See-through projection and display systems

Abstract

Systems and methods related to semi-transparent display screens are described herein. These semi-transparent screens may use horizontal or vertical microlouvers. They allow the display to be placed between a camera and a projector, where the projector is configured to present images on the transparent screen and the camera is configured to record images of a user interacting with the presented images without having to turn his or her back to the camera.

Description

[0001] SEE-THROUGH PROJECTION AND DISPLAY SYSTEMS

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This application claims benefit of and priority to U.S. provisional patent applications: 63 / 698,898 filed September 25, 2025, 63 / 747,555 filed September January 21, 2025, and 63 / 795,233 filed April 26, 2025, the disclosures of which are hereby incorporated herein by reference.

[0004] FIELD OF THE INVENTION

[0005] The invention includes projection and display systems and methods of use, and in particular see-through projection and display systems.

[0006] BRIEF DESCRIPTION OF THE DRAWINGS

[0007] FIG. 1 includes a table showing transparency of louvers as a function of angle, according to various embodiments of the invention.

[0008] FIG. 2 illustrates the data of FIG. 1, according to various embodiments of the invention.

[0009] FIG. 3 illustrates projector offset, according to various embodiments of the invention.

[0010] FIG. 4 includes a table of offset for short throw-lens specifications, according to various embodiments of the invention.

[0011] FIG. 5 includes a table of at various lens offset values at various values for the lens-to-screen distance, according to various embodiments of the invention.

[0012] FIG. 6 illustrates the triangular “wedge” of lens offset that creates a gain in the angularity of the bottommost projected line, according to various embodiments of the invention.

[0013] FIG. 7 illustrates methods of combining additive and subtractive production processes to adapt microlouvers to high-yield production, according to various embodiments.

[0014] FIG. 8A and 8B illustrate interference patterns, according to various embodiments of the invention.

[0015] FIG. 9 illustrates front reflective efficiency enhancement using front treatments of microlouvers, according to various embodiments.

[0016] FIG. 10 illustrates field of view as a function of angular resolution for a 4K camera, according to various embodiments.

[0017] FIG. 11 illustrates a comer of a display, according to various embodiments of the invention. FIG. 12 illustrates left and right reading of a transparent display, according to various embodiments of the invention.

[0018] FIG. 13 illustrates a top view of a transparent display and a user, according to various embodiments of the invention.

[0019] FIG. 14 illustrates data related to MOC can finger-point to positions on literal 36" DM screen, according to various embodiments of tire invention.

[0020] FIG. 15 illustrates data related to Bezel, glass and image dimensions versus screen diagonal measures, according to various embodiments.

[0021] FIG. 16 illustrates a bezel-mounted light source, according to various embodiments of the invention.

[0022] FIG. 17 illustrates an active light module, according to various embodiments of the invention.

[0023] FIG. 18 illustrates a snap-on optical module including mirrors and lenses, according to various embodiments of the invention.

[0024] FIG. 19 illustrates data related to the module of FIG. 18, according to various embodiments of the invention.

[0025] FIG. 20 illustrates a Genlocked HDMI Adapter, according to various embodiments of the invention.

[0026] FIG. 21 illustrates further examples of see-through displays, according to various embodiments of the invention.

[0027] FIG. 22 illustrates angles measured from the plane of the screen, according to various embodiments of the invention.

[0028] DETAILED DESCRIPTION

[0029] Stopping projector glow-through on transparent screens using horizontal microlouvers.

[0030] Louvers at any scale arc parallel slats that arc thin on one axis and broad on a perpendicular axis. With adjustable louvers, changing the angle of the slat determines how much light will pass between slats and how much will be blocked by the broader faces of the slats. Microlouvers are typically not adjustable, so while similar, their impact is not identical.

[0031] Microlouvers are fixed-orientation parallel constructs. Their performance is based on simple geometry: they block light within one range of angles (relative to their placement and orientation while passing light outside that range. In their traditional role, as privacy filters, microlouvers are situated in parallel vertical rows, covering a user’s screen; the user’s view is perpendicular to the screen, so the user can easily see what it displays. People glancing at the user’s screen from either side will not be able to see it.

[0032] FIG. 1 includes a table showing transparency of louvers as a function of angle, according to various embodiments of the invention.

[0033] FIG. 2 illustrates the data of FIG. 1, according to various embodiments. When light comes in at an angle less critical (more perpendicular) than one defined by the geometry of microlouver rows (the green line), it passes through a transparent screen; at sharper angles it’s stopped. Here the louvers are 130 microns deep and 180 microns apart.

[0034] In our context, the parallel lines of microlouvers are horizontal. An ultra-short-throw (UST) projector casting an image upward from below the screen will encounter the louvers from outside their acceptable range of angles. With matte black microlouvers, that light is blocked from crossing the screen either by direct illumination or through reflection or dispersion.

[0035] The table shows those characteristics for a 15-micron wide (0.015 mm or 0.00059 inch) louver of heights from 110-180 microns (0.11-0.18 mm or 0.00433-0.00709 inch) and with separations (gaps) between louvers from 180-190 microns (0.016-0.019 mm or 0.00630-0.00748 inches). The diagram above includes a green line that marks the angle at which more perpendicular light is allowed to pass through the screen while less perpendicular light is stopped by the microlouvers. We chart those angles both as Angle H, against the horizontal plane as a reference point, and as Angle K its angle of incidence related to the vertical plane of the screen.

[0036] FIG. 3 illustrates projector offset, according to various embodiments of the invention.

[0037] Ultra-short-throw (UST) projector lenses are positioned just inches in front of a screen. If that lens is at the same level as the bottom of the screen, the bottom of the projected image might come in at a sufficiently perpendicular (to the plane of the screen) angle to allow it to clear the microlouvers and glow through the glass. In some embodiments, the ultra-short throw projector optionally being configured to be disposed less than 15, 10, or 5 inches from the display service.

[0038] UST projector specifications can include an attribute called offset, which identifies a level for its lens that is below the bottom of the image area. For example, a 120% offset places the lens at 20% of the projected image height below the bottommost level of the image area.

[0039] Attention to the angle between the edge of the lens closest to the screen and the bottom of the image area can inform the selection of microlouver heights and gaps. Enhancing microlouver glow-through protection by controlling ultra-short-throw-lens specifications.

[0040] When using microlouvers on transparent screens as a countermeasure to ultra-short-throw (UST) projector glow-through, the image area of greatest concern is the area closest to the UST lens. Because UST lenses are close to the screen, the projection angle of incidence in that area may be close enough to perpendicular (relative to the plane of the screen) to be allowed to penetrate the microlouver geometry.

[0041] FIG. 4 includes a table of offset for short throw-lens specifications, according to various embodiments of the invention.

[0042] Altering microlouver pass / don’t-pass threshold angles may impact the view from behind the glass (for a camera, for example), but is possible.

[0043] Understanding a small set of simple facts about UST lenses can lead to an easier alternative:

[0044] 1. Most UST projectors are designed to perform a zoom-like adjustment to picture size simply by changing the horizontal distance from lens to screen.

[0045] 2. UST lens specifications include a factor called throw ratio, which is the ratio between the distance of the lens to the screen and the width of the projected image. Many projectors exhibit throw ratios between 0.20 and 0.50.

[0046] 3. A factor called offset allows the level of tire lens to be several inches below the level of the screen. Offset is generally expressed as a percentage of screen height (and often as exceeding 100%). The table at the right shows the size of 10%, 15% and 20% offsets for many screen sizes (in diagonal measure inches).

[0047] Knowing the throw ratio and the image width, it’s easy to determine the lens-to-screen distance.

[0048] Trigonometry (the arctangent of the offset divided by the lens-to-screen distance) gives us than angle by which that combination can lift the bottom of the image upward of arriving near the perpendicular, which would allow glow-through.

[0049] FIG. 5 includes a table of at various lens offset values at various values for the lens-to-screen distance, according to various embodiments of tire invention. Hie table of FIG. 5 looks at various lens offset values at various values for the lens-to-screen distance to show how much each combination boosts the bottom -of-image projection angle above horizontal.

[0050] FIG. 6 illustrates the triangular "wedge" of lens offset that creates a gain in the angularity of the bottommost projected line, according to various embodiments of the invention. Note that as the lens-to- screen distance gets smaller, the angular boost gets larger. Combining additive and subtractive production processes to adapt microlouvers to high-yield production

[0051] Given the significant benefits that microlouver optics can add to products, existing manufacturing techniques are time-consuming and involve considerable manpower-hours, which increases their costs and renders them impractical for many high-volume needs.

[0052] Some of the difficulty in manufacturing them relates directly to their size. Rows of 15 -micron-wide, 130- micron-tall louvers are separated by just 180 microns. Testing shows the challenge of depositing single louver “blades” directly on glass or other transparent media to be insurmountable at this scale.

[0053] FIG. 7 illustrates methods of combining additive and subtractive production processes to adapt microlouvers to high-yield production, according to various embodiments. The process includes:

[0054] The below process solves this in the following steps (referring to FIG. 7):

[0055] 1. Adhesive bonding agent is applied to glass.

[0056] 2. Matte black polymer coating covers surface at desired microlouver depth.

[0057] 3. Lasers, blades or micro-routers remove the coating layer from the gaps between louvers.

[0058] 4. A transparent finishing layer seals the top; note that this layer may also serve other needed functions.

[0059] This method may be used for bonding of microlouvers directly onto glass and other transparent materials.

[0060] Mechanically and optically bonding micro-scale structures and enhancements to transparent (especially glass) screens both as protection and to eliminate sources of unwanted artifacts.

[0061] Step # 4, (See FIG. 7) above, while possibly as thin as 10 microns, caps the glass and microstructures (including microlouvers or front-reflection-efficiency enhancements) In a transparent polymer layer that prevents scratching or mechanical intrusion into these microstructures, blocks entry to dust or particulates and simplifies occasional screen cleaning efforts.

[0062] Avoiding optical artifacts (including moire and interference patterns) in front-reflective glass screens by eliminating layered films when adding such optical enhancements as microlouvers and reflective micro- and nano-structures.

[0063] When light waves encounter spaces in the size range of their wavelengths, or encounter light from different directions, interference patterns can emerge. FIG. 8A and 8B illustrate interference patterns, according to various embodiments of the invention. As illustrated in FIG. 8B, stripe-like patterns can overlap, creating a moire effect. From a web search:

[0064] Both types of unwanted optical artifacts can result from thin polymer films being stacked with each other or bonded to a hard (especially if transparent) surface. Two categories of optical films are likely to be used on display screens.

[0065] One of these categories involves adding vertical microlouvers as a privacy film atop displays in privacysensitive applications, like those under HIPAA regulations orthose, like ATM machines, involving sensitive personal information and access to assets. These films tend to be constructed in multiple layers, wrapping outer ‘’carriage” films around arrays of microlouver chambers.

[0066] The other increasingly common display-related films are so-called “holographic” films, intended to enhance the front reflective efficiency of clear glass or polymer panels. Many of these, too, are multilayer films, including layers intended to reduce the occurrence of the above-depicted artifacts, though not always successful at the task.

[0067] Many such films are delivered with a disposable outer film intended to be removed to expose an adhesive, for bonding the functional film to a hard transparent surface (e.g., glass). While many who order these films attempt applying them by hand, air gaps and wrinkles often occur. This is also a problem area for dedicated film-bonding equipment.

[0068] A production process that directly creates microlouvcrs bonded to glass eliminates the film layers associated with them, and in so doing eliminates a common source of unwanted interference or moire patterns to the assembly.

[0069] This production process can extend to include alternatives to “holographic” films for enhancing front reflective efficiency in an ultra-short-throw projected image environment.

[0070] In the diagram at right, a bonding agent is applied to glass, a polymer coating is applied, and micro-routers, lasers or blades remove “gap” material.

[0071] Between step 3 and step 4, as shown in FIG. 7, there are two approaches that can manifestly improve front reflective efficiency on a permanent basis without including or adding bonded film layers.

[0072] One such approach involves applying a thin layer of a very white coating (like titanium dioxide) to the front edges of the 15-micron-wide microlouvers. Depending on the specifics of the microlouver row separation, this should improve the front reflective efficiency by 23-33% while retaining 91-93% of the transparency of the glass. A table on the next page shows the transparency and reflective efficiency calculations for a range of microlouver heights and gaps (cited in the chart to inform tire included pass / don’t pass light angle calculations).

[0073] The other such approach (which may be performed alone or in conjunction with the one above) adds reflective microstructure or nanostructure materials (e.g., glass, polymer or metal spheres or flakes) to the gaps between microlouver roles just before step 4, above, in which a clear polymer film is added to mechanically protect and stabilize both the microlouver and reflective elements. Typically, the application of all materials, including microlouvers and coating, might increase the thickness of a low-iron glass screen from 3 mm to less than 3.2 mm.

[0074] FIG. 9 illustrates front reflective efficiency enhancement using front treatments of microlouvers, according to various embodiments. Tire data of FIG. 9 shows a range of reflectivity and transparency values likely to result from whitening tire front edges of microlouver arrays.

[0075] Enhancing camera-compatible front reflective efficiency of glass screens by using metallized nanobead sizes smaller than camera pixel perception dimensions

[0076] The benefits of locating a camera behind a transparent display screen are easily compromised if the technology that makes that display image visible imposes visual obstacles into the camera’s view. When the display is created by an ultra-short-throw (UST) projector at its front, a variety of glass, metal and polymer particles may be applied to improve the perceived clarity of the image by improving the screen’s front reflective efficiency. One such reflection improvement approach involves metal-coated glass beads.

[0077] Contemporary camera pixels are microscopic, often on the order of 2 microns square.

[0078] Cameras (like microscopes, telescopes, the human eye or other optical devices) embody several relevant specifications. Those relevant here include:

[0079] • Field of View (FOV), the angle across which detection occurs. Lens choices strongly impact this.

[0080] • Angular Resolution, which may specify, for example, how many pixels are in each degree of FOV

[0081] • 4K UHD Camera Resolution, defined for the dominant standard as 3840 x 2160 pixels (W x H).

[0082] FIG. 10 illustrates field of view as a function of angular resolution for a 4K camera, according to various embodiments.

[0083] The closerthe camera is to the screen, the closer it is to reflection-enhancing particles. Tire table of FIG. 10 places the camera 1 inch from the screen, presumes zero thickness to the screen and, for various FOV (lens angles), how much of the screen, in microns, a single pixel senses. In other embodiments, the camera is 15, 10, 5, or 2 inches from the display surface. Depending on the technology used in the camera sensor, its perception of an object may trigger at less than 100%, even less than 50% of capture. Even if we stipulate that a 25% capture may lead to detection, the range of sizes for the above range of lens angles is 36-115 microns.

[0084] For one product among many for one vendor among several, we note the S3000-S silver-coated glass microspheres from Potters Industries measure 41 microns. Single microspheres of this scale are too small for a camera to perceive. The mathematical exercise represented above informs product selections and can provide practical invisibility for reflection-enhancing products in these circumstances.

[0085] Eliminating unwelcome visible artifacts when placing a video camera close to and behind variously treated transparent front-projection screens

[0086] Transparent (especially glass) screens can be useful devices when ultra-short-throw front projectors cast images on them and are especially beneficial when it is at any endpoint in multiparty communications and user-facing cameras can be placed behind their screens. Attempts to create such devices often encounter unexpected and unwelcome visual / optical artifacts. These tend to fall into classes.

[0087] • Wrong direction - unsurmountable challenges: o Using polarization to eliminate glow-through and suffering a 75% brightness reduction as a result o Drilling or cutting small camera “windows” into covering films only to find the hole edges add a bright circle to the screen o Applying film layers to screens and suffering visible moire and / or interference patterns. o Permitting lighting sources behind the screen to feed reflections from the back of the screen into the camera.

[0088] • Flawed approaches and incomplete treatments o Placing both camera and projector behind the screen, resulting in glare and reflections in the camera image and harsh brightness plus incomplete displays for the user. o Underestimating the brightness needed for vivid displays in normal office ambient lighting o Failing to account for glass edge-effect internal reflections

[0089] • Bad starts o Adding “holographic” front-reflective-efficiency films without using bonding fixtures o Adding uneven distributions of reflectivity-enhancing products o Compromising transparency with too-heavy distributions of reflectivity booster products o Placing privacy films between camera and user and seeing a rippled forehead effect. FIG. 11 illustrates a comer of a display, according to various embodiments of the invention. Even a small bezel blocks ambient light entry at glass edges. To prevent projector light internally reflecting from glass edges into moire or interference patterns, a gap between the inner edge of the bezel and the outer edge of the image area is usefid. Both internal glass reflections invite unwanted artifacts.

[0090] Just as the glass screen can reroute light in often unexpected ways, the camera lens can detect or respond to unexpected light sources.

[0091] In any environment with any active projector, random reflections may direct that light into the camera lens, causing glare. In any environment where ambient light is maintained, stray light may arrive at the lens and be enhanced by internal reflections within the lens, an effect called lens flare.

[0092] Traditionally, these can be prevented by adding a lens hood - usually a round or rectangular shape that fits just over the lens at the camera and then expands outward to always be larger than the field of view of the lens. It blocks light from all lateral sides.

[0093] But the lens hood itself can introduce a challenge. Most lens hoods are matte black, inside and out, to absorb rather than reflect light. When a camera behind a transparent screen has an all-black lens hood attached, it’s perceived by the user as a dark spot, disruptive of the projected image.

[0094] Making the inside of the lens hood something other than black is not necessarily a satisfactory solution. Any coloration within the lens hood would be perceived as a tinted area of the projected image. Too white (or too reflective) a treatment could capture light from the front of the lens and create a reflection on the back of the screen that the camera will see.

[0095] This visible artifact is best eliminated by choosing a very matte, very light gray color that even in the presence of a directed light source from the front of the screen would produce only a very muted reflection on the back.

[0096] Eliminating glass (or other medium) edge reflection artifacts in designing and deploying transparent front-projection screens

[0097] FIG. 11 illustrates a comer of a display, according to various embodiments of the invention. This comer includes a gap between tire display surface and a bezel surrounding the display surface.

[0098] Optical design criteria for embracing finger-pointed image references in front-reflective projected displays on transparent screens without requiring digitization through electronic digitization or artificial-intelligence interpretation. In professional video studios (e.g., broadcast weather or election reporting), using a transparent glass screen displaying graphics from an ultra-short-throw (UST) projector - and placing a studio-quality camera behind the screen - means a presenter never has to turn his / her back to the camera to be involved with the graphics. Those graphics can be electronically composited with video of the presenter to appear to be on a semitransparent plane between the presenter and the camera.

[0099] FIG. 12 illustrates left and right reading of a transparent display, according to various embodiments of the invention.

[0100] The presenter sees the projected graphic as right-reading. The camera’s view ofthe presenter can get mirror- flipped so that the composite image is also right-reading to the audience.

[0101] But the field of view of a camera behind a transparent screen does not necessarily reach the full extent of the image projection area on the glass - but here, it must - or the fingertips of a presenter pointing to an outlying position would be outside tire area tire camera can capture.

[0102] There are challenges in finding the best balance of camera placement behind the screen, field of view so the image covers the entire display area (but nothing beyond), and some concern for how that impacts the camera’s view of the presenter (at arm’s length in front of the screen) and his / her surroundings.

[0103] For the top-view diagram that follows - based on one selected set of data (a broader table follows), our example uses the following specifications:

[0104] • 36” diagonal-measure 16:9 aspect ratio screen with a 31.4” image width

[0105] • Camera with a 48° FOV (field of view) lens placed 35.3” behind the screen

[0106] • The presenter 18” behind the screen but a focal point at 19” (more on the eyes than on the nose); note that the depth-of-field of the camera will keep the presenter’s torso well in focus.

[0107] • Assuming a 24” torso width, the presenter will occupy 51% of the screen width

[0108] Tire camera's distance from the screen also means that 15 -micron-wide microlouvers will be invisible to it.

[0109] The following table also follows the example use of a 36” diagonal-measure screen and the presenter 18” in front of the screen. To fit the camera’s framing to the projected image size, it looks at a range of camera field-of-view (FOV) lens angles from 35°-60°. For each of these lens angles, it shows the required camera- to-screen distance to accomplish matched framing.

[0110] The blue column shows the total width the camera sees as of the position of the presenter. The green column shows the percentage of that image width that a 24”-wide torso will occupy. And it calculates the net focal length from the camera lens to the presenter’s eyes. To great extent, the choice among these is a matter of preference or of circumstance. Where studio space is tight or ’‘clean” background (behind the presenter) space is limited, the blue column may get a priority. Where space behind the screen is at a premium, the camera-to-screen distance may prevail.

[0111] Tire example’s indicated selection of a 48° field of view considered the 51% presenter image width and a camera position less than a yard behind the screen.

[0112] FIG. 13 illustrates a top view of a transparent display and a user, according to various embodiments of the invention.

[0113] In sum, these criteria apply to letting a real person’s real fingers point to any position on the screen that viewers see and are necessary to allow that.

[0114] FIG. 14 illustrates data related to MOC can finger-point to positions on literal 36" DM screen, according to various embodiments of the invention.

[0115] For broadcast applications involving MOC pointing at on-screen positions, we can capture the full screen area (but none of the bezel). The capture width at the MOC will be greater; assuming (arm's length) 18" from to screen and 24" torso width, MOC % of image width is shown in green column.

[0116] Methods for minimizing the visual impact (perceptibility) of a bezel surrounding a transparent screen.

[0117] Most displays are mounted in surrounding bezels for any of a variety of reasons. Some examples:

[0118] • Prevent delamination of bonded screen layers

[0119] • Prevent edge damage to glass

[0120] • Disguise wire routing

[0121] • Subtly embed w ebcams or other features

[0122] • Provide a handhold area that doesn’t contribute to screen fingerprints

[0123] • Provide a useful handhold w hen mounting

[0124] • Prevent light from entering or exiting the edges of screen glass

[0125] Most (not all) display screen bezels are black, and their presence may have been inspired by picture frames. Visually, most bezels tag human space perceptions with an outline, drawing attention to the product within.

[0126] Transparent screens present a different set of priorities. Many familiar images of transparent screens from futuristic film or television fiction represent them as unframed panels of glass. And thereby, for practical transparent screens that may require a bezel, there may be compelling reasons to render it much less obvious than a black outline.

[0127] We will use a selected practical example in our discussions - a 36” diagonal measure 16:9 display (17.6” inches tall and 31.4” wide) that uses a glass screen 3.1 mm thick.

[0128] Tire following methods can help make the sight of a properly designed bezel a visual non-event:

[0129] SIZE: The glass is 3.1mm thick, so without a bezel, that’s the size of its edge that people would see. A bezel might be extruded aluminum with a very thin (0.020”) wall thickness, making for a maximum bezel thickness of 4.5 mm. just 50% bigger - and. on that scale, difficult to scale. The sides of the bezel that hug the glass are 4 mm long - roughly5 / 32” . The height of the screen is 17.6 inches (roughly 450 mm), more than a hundred times taller than the bezel. There are critical mechanical requirements for the bezel at these sizes, but those are not part of our subject here.

[0130] COLOR: The edge of highly transparent glass, like low -iron glass, does not have the familiar green tint of home improvement store window panes. Its appearance is close to white (hex FFFFFF). A light gray coloration (like DBDBDD) would be more subtle.

[0131] CONTOUR: While smaller sizes may help escape notice, and more muted colors also assist, there are other attributes of a mechanical structure like a bezel that also tend to draw the eye. Those include hard comers, flat surfaces and step-gaps between the edge of the bezel and the edge of the glass. The methods to invoke here include very minimally outward-rounded flat surfaces, comer rounding with a radius not much larger than the metal thickness, a flat front extent where it meets the glass (to help it visually blend) and a very tight front fit so the edge of the bezel where it touches the glass is difficult to discern.

[0132] INTEGRATION: On a more macro scale, the minimal reach of the sides of the bezel needs attention to prevention of unintended elasticity; shifting or twisting. (One image to consider is using a large rubber grommet or rubber band - what would keep it snug?) The bezel and glass must integrate into a frame. Consider a bezel along the top and 3 sides of a glass screen . . . a beefier base beneath... and downward extensions of the bezel secured to the base. The glass should also slide into the base. A tiny foam gasket surrounding the glass and pocketed within the bezel and the frame helps complete the integration by unitizing the structure using the elastic compression of the gasket.

[0133] FIG. 1 illustrates data related to Bezel, glass and image dimensions versus screen diagonal measures, according to various embodiments.

[0134] Bezel / frame-mounted beamed illuminators Contemporary examples of user lighting by illuminators built into webcams or into display frames include only “wash” or diffuse lighting, which tend not to produce flattering results.

[0135] Professional film and video lighting techniques use beamed illumination, directed and shaped to perform specific tasks. In combination, these combine to provide good illumination to tire eyes... to the area that a person’s hand covers when tire heel of the hand is on the chin, the palm over tire nose and the fingertips at the eyebrows. . . and. from both sides, the contours across the cheeks.

[0136] Like studio lighting, facial illumination from a frame- or bezel-mounted light source must be shaped and directed. Like studio lighting, the illumination setup should be determined in the absence of the user and not be subject to being misadjusted by a user. Like studio lighting, the color rendition of the lighting should be true to normal color perceptions, an especially tricky factor when fabrics or makeup are present. Like studio lighting, there should be no partem of dots, spots or stripes cast by the lighting, and no flicker.

[0137] Because the angles to tire targeted facial features are different for each display frame size (a table follows), we will suggest a 2-piece approach. A module that mounts to the outside of the bezel contains the active electronics, the lensed LED illuminators and 45°-angled mirrors to focus a long beam perpendicularly away from the bezel. The LED emitters have lenses that are selected to avoid projecting uneven light. A snap-on optical module redirects that beam at an inward angle that’s specific to each display size.

[0138] FIG. 16 illustrates a bezel-mounted light source, according to various embodiments of the invention.

[0139] In our test bed system, all emitters arc selected at a color temperature of 4100° Kelvin, each of 3 strips (for the center of the top, left and right bezels) produces roughly 400 lumens for a total of 1200 lumens, meaning the equivalent brightness corresponds to a 100-Watt bulb. They tap a 12 VDC regulated power feed inside the bezel and connect through a bridge, so there’s no such thing as a reversed connection.

[0140] FIG. 17 illustrates an active light module, according to various embodiments of the invention.

[0141] This sketch depicts the makeup of the active module. The black at the bottom depicts the front face of the bezel. A PC board with the driver components attaches to tire bezel and connects to a DC power feed within. A perpendicular (to the bezel face) array of lensed emitters (the sketch does not represent the actual layout of these) beams light to 45° -angled mirrors and as the ray diagram indicates, we get a controlled, forw ardpointing beam.

[0142] FIG. 18 illustrates a snap-on optical module including mirrors and lenses, according to various embodiments of the invention. The active module is designed to be uniform across all screen sizes, but to accept a passive module that can snap onto it. This diagram shows the mounting of the active module on the bezel face and how the sleeve of the passive module fits over and snaps into the active module.

[0143] These diagrams represent cross-sections of the lighting modules. Depending on the selection of LED emitters, each is likely to be an elongated strip, perhaps 7 inches long.

[0144] There are many optical mechanisms that, in the passive module, can redirect perpendicular light to a new angle. In many circumstances, simple mirrors can work.

[0145] The chart on the next page looks at display sizes from 24 to 66 inches diagonal measure, and for each, shows the width and height of the bezels we use.

[0146] Stipulating an arm’s-length 18-inch user distance to the screen, we chart the distance from light to face from the top and from the sides and calculate the offset angle to which the passive module should repoint the light.

[0147] When facing a camera, ambient lighting alone tends to be diffused, flat and at unpredictable color temperatures, which can impact the perception of colors in clothing and makeup. Accessory lights (like ring lights) may be better than ambient light alone, but they do little to assist the perceived contours, depth and dimensional characteristics of a face. The level of light these beamed illuminators can add, in addition to providing a more flattering personal appearance, also increase the illumination “floor” at the camera, potentially raising it to a level where subtleties become easier to share and picture noise ceases to be a factor.

[0148] FIG. 19 illustrates data related to the module of FIG. 18, according to various embodiments of the invention.

[0149] An add-on device for HDMI video connections to lock frame initiation to a studio genlock sync signal.

[0150] In professional video studios, it’s important for cameras and displays to all “march in step” to a locally generated frame synchronization signal. The signal connectors tend to be BNC coax fittings labeled “SYNC” (sometimes with “IN” or “OUT” appended. The practice of coordinating everything through the locally generated signal is called “genlock”.

[0151] When cameras, displays and other equipment is not in sync, the results can be sloppy, unpleasant and sometimes unusable. You might see rolling, tearing or other unwelcome corruption to viewing. When professional video studios connect video signals to various midpoints or endpoints within the studio, it is usually over a coaxial cable using the SDI (Serial Data Interface) protocol. SDI data is roughly equivalent to the data carried over HDMI cabling, except that SDI takes, broadly, a serial approach and HDMI takes, broadly, a parallel approach.

[0152] While most SDI-connec6ted equipment comes equipped with a SYNC connection, most HDMI equipment does not.

[0153] Most SYNC signals take one of two approaches. The elder of the two is an analog "‘Black Burst” sent at the beginning of each frame: the newer is Tri-Level Sync or Tri-Sync. It is possible for a single input to read and react to whichever standard is present.

[0154] FIG. 20 illustrates a Genlocked HDMI Adapter, according to various embodiments of the invention.

[0155] The GENLOCKED HDMI ADAPTER is an adapter that reads an incoming SDI video feed and a SYNC input and feeds an output HDMI connection with a video signal that’s genlocked to the incoming studio sync connection.

[0156] It essentially provides HDMI-connected display equipment with tire SYNC input they don’t have to make them compatible with genlocked production environments.

[0157] Latency elimination in systems that can composite synthetic images of added graphics with video camera images of its user.

[0158] The act of compositing multiple sources into a single video image has been commonplace since the 1960s. In that era, it most often appeared as a “super”, short for “superimposition”, where frames w ould overlay each other. In modem video production, digitization has added options and made the process much easier to perform and to support.

[0159] Current virtual meeting software does not make it easy to composite the local camera’s vision of a user with graphics that the user may wish to also display.

[0160] For purpose-designed equipment that, for example, projects a video image onto a transparent screen while placing a camera behind the screen that is always “eye to eye” with tire user, image compositing offers significant boosts to communications. When the images of the user and the graphics are joined, the graphics can appear to be on a floating semi-transparent layer between the user and the camera.

[0161] FIG. 21 illustrates further examples of see-through displays, according to various embodiments of the invention. The combined (overlayed) image does not exist in the real world; it’s electronically synthesized. Most attempts to perform such compositing rely on an SoC (System on a chip) which incorporates a CPU and, optionally, memory, communications and other functionality.

[0162] An SoC is, in many way, a computer, and like a computer, needs to have software written for it, firmware written to it and so on. An SoC can be an inexpensive way to perform the video compositing being addressed here. The choice of a specific SoC will often consider how many video streams, at what resolution and which frame rates must be handled.

[0163] Like most computers, the SoC runs its programs one command at a time. At 60 frames per second, one frame of video lasts 17.67 milliseconds. The compositing function, even on a sufficiently fast SoC, can introduce latency (meaning delay) at three points during its processing. This latency can amount to, then, a total of 1 / 20 second. That may not seem like much as a printed number, but when you're watching video it’s enough to let you notice that lips are not keeping up with the audio they produce.

[0164] One-command-at-a-time operations are also called serial operations. Tire SoC is a serial device.

[0165] Where serial is a sequential approach, its alternative takes a parallel approach. The FPGA (for Field Programmable Gate Array) is a parallel device, combining actions in hardware based on its programming. Where the SoC programming is a list of chores to perform, FPGA programming is a description of the configuration “blueprint” for many hardware gate devices, all at one time. (This is an admitted oversimplification, but not misleading).

[0166] Eliminating latency when compositing two or more content sources in real time means a commitment to FPGA hardware that many developers are reluctant to take. But simply, the latency introduced by SoC processing of multiple video sources will always be present and consequentially compromise the resulting video, while basing systems on adequately powerful FPGA cores introduces latency that is so minimal that not a single line or frame gets dropped.

[0167] Combining angular targeting with range limitations when deploying PIR sensors to frontprojection screens to extend the period of viability of a projector’s illuminator(s).

[0168] 3-laser RGB projectors with illumination levels under 8.000 lumens depend on lasers rated for a life expectancy of 20,000 hours, which amounts to roughly 2! years if run 24 / 7. A user who treats a transparent screen as a second monitor and who is accustomed to running a second monitor 24 / 7 can require replacing the lasers (a complex process) as soon as 2+ years. This discourages projector manufacturers and manufacturers of systems that embed projectors from offering 5-year warranties, which are popular among buyers. PIR-based motion detection is commonly applied to such circumstances, but this is an approach that often falls out of favor with users.

[0169] There is a combination of factors that make contemporary PIR products a better choice in the above application than predecessor PIR products would have allowed.

[0170] Two new factors add sophistication to PIR choices: the ability to limit their detection range and the availability of a broader choice of purpose-built lenses with defined fields of view.

[0171] For a display design with UST front projection onto an enhanced-reflectivity transparent screen mounted on an extruded aluminum base, a selected PIR device / lens combination can occupy lower extremes of such a base, perhaps %” above the bottom of the base and %” from the left and right edges.

[0172] The expectation of a user being within arm’s length (18”) of the screen allows straightforward calculations of the best aiming angles to detect a user’s torso motion. Given that proximity, it’s possible to require both PIR's detecting motion (and potentially to detect when there are repeated instances of only one reporting motion, which could suggest a need to replace the other). Whether using an either or both trigger strategy, the nominal use of such transparent-screen products centers on virtual meetings.

[0173] NOTE: If a significant tactical goal of the motion-detection technology is to extend laser life to 5 full years, that means a (365 -day) daily average of no more than 11 hours of use per day (no more than 15 hours for people who take weekends and holidays off). With a generous 90-minute wait before a lack of motion turns off the projector and an average 8-hour work day, the lasers should still be viable after more than 5 years.

[0174] The following chart covers the PIR pointing geometry’ for displays with diagonal measures from 24 to 65 inches. It shows beam angles for both sides to the closest edge of a torso, to its center and to a mid-point. It shows the vertical angle to the user’s eye level. And it shows the required sensor field of view (FOV) for detection.

[0175] NOTE: The reason for range-limiting is to prevent background motion from impersonating user activity. In environments where, for example, cleaning crews come through eve ’ night, the stay- on time can be shortened from 90 minutes to something less.

[0176] FIG. 22 illustrates angles measured from tire plane of the screen, according to various embodiments of the invention.

[0177] If most past disappointments regarding PIR-sensor-controlled services like lighting were based on sparse control over where those sensors were looking, this approach has usefulness well beyond projectors. BRIEFING: Glass as screen.

[0178] When any projector casts an image onto a transparent surface, the visibility of the image to an observer, as well as its presence in the context of the space surrounding the screen, can be corrupted or disrupted by simple science:

[0179] • A transparent screen normally lets light from the projector shine through, and the resulting offscreen images can detract from the intended display. (This effect is often called glow-through).

[0180] • The screen can directly reflect an image of the projector’s lens - during projection a very bright light source - which appears as a bright spot or circle on the screen, making it very difficult to see the intended projected image.

[0181] • The more transparent the screen, the less it reflects light and thereby the less vivid the observed projected image.

[0182] BRIEFING: Traditional approaches to increasing front-projection reflective efficiency

[0183] • Screens can improve the vividness of a front-projected image by adding translucent to opaque reflective elements, but these impose on the dramatic effect of screen transparency and reduce the visibility of objects on the far side of the screen (the visual incorporation of which is often the express purpose of the image projection effort.

[0184] • When front-reflective-efficiency-improving treatments are applied directly to the screen, their distribution is often uneven, resulting in a projected image that may partially be very vivid while partially being more obscure.

[0185] • Any beyond minimal distribution of front-reflective-efficiency-improving elements or treatments on the screen reduce the perceived (and real) effective transparency of the screen.

[0186] • Many modem front-projection reflective-efficiency -enhancing products are called "‘holographic”, though there seems to be no volumetric enhancement to their treatment of light. These tend to be embedded in single- or multiple-layer polymer films, which memorialize a controlled distribution of reflective materials. The distance from such films to the screen surface (or between their own layers) may be within the range of wavelengths of light, resulting in such unwelcome attributes as moire or interference patterns.

[0187] BRIEFING: Microlouver privacy filters as glow-through countermeasures

[0188] Like a microscopic implementation of window blinds, privacy filters using parallel filaments as microlouvers permit straight-through light flow but block light that comes at an angle. • In privacy mode, the microlouvers run top to bottom and mask views of the screens behind them from either side. One 3M executive explained it as intending to prohibit an airplane passenger in a seat on either side from seeing a passenger’s screen.

[0189] • The patent on microlouvers as privacy filters has expired.

[0190] • Modem commercial products using microlouvers are available embedded between layers of film, and as a result can introduce unwelcome moire or interference pattern attributes when applied.

[0191] The term “microlouver” is justified by the truly microscopic dimensions of their louvers. Precise measurements of a popular contemporary product showed the following: Lower edge width: 15 microns, Louver height: 130-150 micros, and Louver-to-louver spacing: 180 micros.

[0192] PRIMARY AND SECONDARY ROLES

[0193] Embodiments of the various inventions discussed herein are solutions to at least one specific set of circumstances that also apply and will be important to others.

[0194] At its core, the challenge involves creating a transparent screen on which an appropriately configured projector can cleanly and vividly create a 2-dimensional display. Moreover, it should permit a camera mounted behind the display a clear and potentially enhanced view of a user in front of tire display.

[0195] Doing so requires solving optical, visual, lighting and other challenges. These elements handicap the ability of a simple pane of glass (or other transparent material) to fulfill this core commitment:

[0196] • One-way transparency is not native to any single material and has not been done without introducing major compromises to that transparency. Efforts to accomplish it have tended to use opposing linear or circular polarizers at both the reflection and the screen, which at once reduces the projector brightness by 50% and the screen transparency by 50%.

[0197] • Opaque projection screens seek perfect reflectivity (reflective efficiency) by enhancing native white or metallic front surfaces with reflectivity boosting treatments like glass or metallic particles. Transparent screens may reflect a bright image of a projector’s lens but are natively inefficient at reflecting images. Glass or metal particles or polymer coating or flake treatments can improve their reflective efficiency, but tire greater the improvement the less transparent (and the more opaque or translucent) the screen.

[0198] • The less reflective the screen, the larger, more powerful and more expensive a projector must be. A professional rule of thumb says the brightness of a projected image reflecting from a screen (as measured in Lux, for example) must be 5 times the location’s ambient brightness. As an example, office spaces with typical ambient light levels of 1000-2000 Lux would need to create an image that emits 5000-6000 lumens from the screen, but that’s at 100% reflective efficiency. Highly transparent screens today tend to have a front-projection reflective efficiency of 5-25%, requiring projectors capable of producing 20,000-120,000 Lux.

[0199] Tire primary role of equipment to which these designs apply involves transparent displays with cameras behind the screen, intended for use in virtual meetings, and for live or recorded video production.

[0200] Secondary roles include displays as destinations (for example, in museums) or as attractions (for example, on store windows or for themed events). Our methodologies also allow the production of privacy shields at dramatically lower costs than current methods.

[0201] Software able to derive a single schedule of upcoming meetings from a variety of applications can address a commonplace challenge that results in missed meetings and multiple simultaneously scheduled meetings booked in otherwise disconnected sources.

[0202] Given tire variety of ways in which scheduling can happen for both online and physical meetings, there is no known single venue that coordinates all such schedules without meticulous attention from individuals. Given the vagaries of the various scheduling entities, there is no single software approach for determining all such scheduled meetings.

[0203] As a result, the approach described here provides a unique solution to this challenge. It involves the integration of multiple schedule-harvesting techniques into a single software operation that may be run manually by an individual or automatically as a background activity.

[0204] D ATA-HARVESTING VECTORS

[0205] The following techniques are incorporated, and other techniques are possible:

[0206] • FRIENDLY API INQUIRIES: Vendor / creator-published application programming interface (API) elements, where available, may allow extraction of scheduled meeting infonuation from specific online meeting, scheduling or calendar applications or utilities, and possibly from email applications that carry meeting invitations and invitation responses.

[0207] • UNDOCUMENTED APPLICATION INQUIRIES: Without disrupting, corrupting or otherwise endangering the data stores associated with the classes of software applications and utilities cited above, it is possible in many cases to examine data within their direct data stores or within “snapshot” (temporary local) copies of those data stores to extract meeting schedule information.

[0208] • AUTOMATED UX OPERATION PLUS AUTOMATED RECOGNITION: Software can mimic user interface (UX) operator commands specific to each such application to bring up screens that display scheduled-meeting specifics, and the pixel patterns within those screens can be analyzed with recognition algorithms to convert human-readable text into machine-readable text. This can be made available as a user-initiated uti 1 i ty that may optionally allow7user opportunities for editing or corrections, or automatically and unobtrusively through the software creation of an unseen virtual screen.

[0209] SCHEDULE INTEGRATION, COORDINATION, ALERTS TO NEEDED ALTERATION

[0210] The specific information included in any one application’s (or utility’s) record or description of a meeting is unlikely to represent a one-on-one match to the format for information stores used by any other application or even to future versions of the same application.

[0211] Accomplishing integration is best approached by maintaining a super-set of every application’s individual items and their formatting (in database tenns, field mapping); this also requires maintaining a reference “dictionary” of field maps for each such application. This core data asset is a significant help both when ingesting schedule data and when updating information in such places as a master calendar.

[0212] Translating (meaning, here, algorithmically modifying information from one source for clarity when output to a different source) will most likely come into play when providing information about all scheduled meetings to the user’s (or organization’s) favored choice among calendar or personal scheduling applications or utilities, ft allow s including the identification of which online meeting application or which local meeting facility will be the destination for participants. It also provides a mechanism for imposing some standardization to secondary attributes of the display of scheduled meetings; for example, the identification of a category or the imposition of a color code or the setting of how long before an event to provide a reminder.

[0213] It makes sense to anticipate the likelihood of scheduling conflicts, especially between meetings scheduled by different applications. Users can choose among available alerting mechanisms; these might include calendar alerts, text messages, on-screen message boxes, among other options.

[0214] PERIPHERAL PRODUCT INTEGRATION

[0215] When the addition of a peripheral product relevant to online meetings brings with it supporting application or utility software, integration of this more universal awareness of scheduled meetings can extend the convenience nature of such products. It could, for example, at the end of any given online meeting, display the time and venue for the next scheduled meeting across all online and local alternatives.

[0216] Publishing or export of aggregated schedules originating from multiple application sources to one or more selected applications, utilities or other software or display assets. Workplace (or personal) productivity may suffer when end users must consult multiple resources simply to have certainty about their time commitments within any span of time of interest. As little as 2 minutes per scheduling source and 1 minute per item information transfer to a single “master” (personally favored) scheduling application or utility can cost 15-30 minutes of time that is, from any external perspective, unproductive.

[0217] Once such productivity drains are recognized, they will (and must) inevitably lead to methods or mechanisms that aggregate schedule-related data from a variety of sources. These may. for example, include online or physical meeting invitations or notifications, entries into calendar or scheduling utilities, or schedule-related messages in email messages. The least unlikely data handling approach expected from such aggregators involves their own data stores of collected infonnation.

[0218] Any such aggregator intended to reduce the need to consult multiple application or utility sources must necessarily introduce its own time burden if it intends to become “yet one more” resource for schedulechecking, offering only its own user interface for access to tire aggregated scheduling information.

[0219] A more productive solution, whether integrated into such an aggregator product or performed externally to it, is to read the data store of such an aggregator, to format its content for clarity, and to forward that content through an acceptable method (for example, export, publication, posting or other integration ingest) to one or more user-selected destinations.

[0220] Such destinations may include a user-favored calendar application (like Outlook), a calendar-printing application or any preferred venue for presenting a convenient visualization of schedules. Further, for managers, such export or publishing facilities might offer a selection of formatting choices for visualizing the scheduling of multiple managed people (and / or facilities, like conference rooms).

[0221] Presentation of source-agnostic next-scheduled event information at the conclusion of a virtual and / or practical meeting.

[0222] Scheduled events are escalating occurrences for both businesses and individuals in non-business contexts, as well as for many facilities that require reservations. Hie common thread across all these potential calendar or agenda entries is that they each involve a specific date (or multiple dates, where recurrence is a factor) and a specific start and end time. The uncommon denominator is the source of such scheduling.

[0223] Requested or required participation in virtual meetings may originate in many ways, including meetings scheduled by any of a variety of openly available or internally specialized virtual meeting software entities (applications, utilities or online facilities), or meetings scheduled through desktop or online calendars, schedules, to-do-lists or other applications, utilities or online facilities, or meetings scheduled through email or other electronic (or printed) communications. Similarly, the scheduling of facilities may originate from one or more sources.

[0224] Regardless of the single or multiple methodologies through which source-agnostic scheduled event infomration may be aggregated, it is most useful when easily available to its user(s) or other relevant people.

[0225] Two observed office and remote workplace phenomena suggest counterproductive outcomes: when a relevant person forgets or overlooks the time or venue (e.g., specific virtual meeting software application or utility, or specific physical meeting room) for an event, especially a scheduled event to which confirmation was sent.. . and when a meeting time is not forgotten or ignored, the attempt to join the meeting is wrongly associated with the wrong choice of virtual meeting application or utility.

[0226] Noting the research identifying that escalating involvement in online meetings introduces fatigue and distraction, it is reasonable to conclude that next-event information - at a minimum, including both the starting time and the venue for the event - enjoys heightened usefulness when displayed at the conclusion of one or more previous scheduled events.

[0227] The scope of aggregation of event schedules and venues may be as broad or as narrow as any specific circumstance mandates, and may include any combination of virtual or physical venues or facilities or assets; regardless of the scope of that aggregation, immediate reminders upon the conclusion of any one event of relevant next-scheduled event(s) may help prevent diminished awareness of the information disseminated during missed events, and in so doing, increase productivity and reduce stress in the workplace.

[0228] Using coarse object-recognition Al to recognize whether a hand or a head dominates a current full or partial display view.

[0229] Since childhood, people learn to raise their hands to elicit attention from a supervisor or a leader. While this is a more orderly practice than simply shouting out statements or questions, the practice offers no immunity from unproductive interruptions proffered under the “flag” of the raised hand. Tire practice of raising hands is also a common mechanism for detennining, in a real or metaphorical sense, by vote, the extent to which those who are (virtually or physically) present favor specific choices within a range of alternatives.

[0230] Online (virtual) meetings represent one common contemporary example of raised hands becoming a vehicle for interruptions that are often harmfill to the interests of those initiating the activity. Interruptions in this context can make it more difficult for presenters to efficiently present and for participants to readily and clearly understand the content being communicated. Such interruptions can also be reasons for meetings running longer than necessary, for not completing planned agendas for a meeting and for reducing participant focus (often by diverting it to other chores) when tangents or diversions are not of general interest; research shows this to be a factor in worker fatigue, in reduced creativity and, for the organizer, in a reduced productivity return on their investment in payroll. Despite these identifiably negative impacts, some virtual meeting software offers a “hand-raising” user interface element.

[0231] Within the Al (artificial intelligence) “toolbox”, machine learning provides a reliable way to analyze images in useful ways. In broad terms, it allows computer analysis of a very large number of images that include examples of single or alternative elements of interest as well as images that include none. In a sense, machine learning algorithms, while initially guided by human “score-keepers”, eventually train themselves to reliably identify which among several elements is present, or if none of the alternatives is recognized.

[0232] Within the example context of virtual meetings, the connected camera (even if its image is not shared with the meeting) is either optically disabled (using a shutter or cover, for example) or points at the user's head and upper torso. Machine-learning can recognize a head within such an image; the width difference in the transition from head to shoulders simplifies that task, as does the approximately oval shape of the head. (Note: there is no need to accomplish or attempt facial recognition in the process of recognizing human heads).

[0233] Presuming that the user is able and willing to follow simple instructions (e.g., spread your fingers and extend your ami to point your palm toward the camera), the more complex (but still predictable, even for hands offering fewer than 5 digits) shape of a hand is also easy to recognize. Even if the spread of fingers still allows the camerato see the shape of ahead behind the outstretched hand, machine learning can provide recognition of the presence of the hand.

[0234] For most “raised hand” applications, the recognition task involves only two choices - a yes or no - to signal whether the presence of a raised hand has been detected. When this facility is implemented, individuals or organizations can elect to selectively disallow raised-hand recognition unless a vote-style show-of-hands is requested, or only during a specified period that invites questions, comments or other interruptions; these practices may also be retroactively imposed on raised-hand user-interface elements, of course, but automated recognition may prove to be especially advantageous where future meeting or meeting- adjunctive applications can better exploit it.

[0235] Applying hand versus head recognition to automatically track individual “show of hands” counts among the individual participants in virtual meetings.

[0236] Tire ability to recognize when a raised hand is a foremost element of live video from a connected computer (e.g., during online remote meetings) provides an opportunity to collect either absolute or relative counts that may be productive in several ways. This describes methodologies for applying such a facility and describes their benefits in comparison to traditional techniques.

[0237] Limitations on the applicability of an automated “show of hands” count depend on the ability of the virtual meeting’s hosting software to display video images of all participants, on the total number of participants (and whether it exceeds the aforementioned limit), on the willingness of participants to display camera video during a meeting (even if only during a call for a “show of hands”) and on the actual presence of the participant at the connected computer through which he or she is assumptively (but not necessarily) a meeting participant.

[0238] In the context of virtual meetings, relative counts exist at two levels, based on empirical observations that indicate a strong likelihood of incomplete participation in any activity that calls for a “show of hands” count. One of those levels is the balance of counts against each other, whether a limited one-or-the-other (or pro / con, for / against or guilty / not guilty) selection or a broader one (e.g.. which one of these several candidates should speak on our behalf). Only the “show of hands” of meeting participants who do show hands can be counted and their number may represent less than the full list of connected participants in the meeting.

[0239] At one level, relative counts may refer to the total number of “show of hands” responses gathered without regard for abstentions, whether deliberate or circumstantial. At another level, relative counts may factor in the total number of connected meeting participants. A third option represents a hybrid approach, where those among the participants who abstain from voting are separately calculated.

[0240] Some circumstances may call for absolute counts. Consider, for example, conditions in which something needs to be bought, e.g., how many of you wear a size XL T-shirt (and counts for other sizes)? Who would like us to provide a vegan entree? Will you need a ride?

[0241] In any of the above cases, and within tire stated limitations, automated “show of hands” counting within groups of people already gathered for a virtual meeting saves time, improves accuracy and infuses fewer distractions than other methods of determining similar outcomes.

[0242] Using real-time Al to recognize direct versus diverted eye contact.

[0243] Video-connected communications venues through computers, tablets, smart phones or other technologies largely make use of video cameras that are either embedded in a device bezel, embedded near the edge of a user-facing screen or mounted atop a display bezel. This complicates the task of mining a real-time video feed of a user's face to determine whether the user is making and maintaining direct eye contact with onscreen content or diverting eye contact elsewhere. Because eye contact is a primary behavioral indicator of attention, feedback regarding the extent of direct eye contact versus diverted eye contact represents valuable information to the presenters, producers, sponsors or managers of such video-connected communications. In this regard, the category of video- connected communications includes one-on-one video calls and virtual meetings as well as other applications.

[0244] Direct eye contact with a camera in its simplest form represents a view from the camera of an individual’s face with the horizontal center line of both pupils equidistant from the centerline of tire nose (on glabella bone at the nasion), where tire iris of each eye appears centered in the white (sclera) of that eye. This is a rare circumstance because it would require the center-point of the video screen to coincide with the centerpoint of the camera lens and the two center-lines to coincide.

[0245] The actual eye-line of a user does tend to be consistent with the center-point of the user’s video display, but the lens axis of the associated camera that captures the image of the user’s face is off-center, near the edge of the display for smart phones or some tablets, in the bezel for tablets and some video monitors, or in a separate camera mounted on the bezel of many video monitors. From the perspective of the camera, as seen in the video it captures and thereby in the video of the user as seen by the meeting, eye contact will be seen and perhaps assessed as diverted; research shows that most people associate this with evasion.

[0246] Note that the cluster of off-center positions a camera may occupy, while not totally predictable, tends to include a relatively limited number of possibilities. When the user’s eye contact is diverted away from the screen, it’s reasonable to assume that many to most such focal points will fall outside the more limited likely positions of integrated cameras.

[0247] Even without assistance, machine learning against a very large number of sample images (or videos) can soon detect the repeatedly observed cluster of landing points that represent the user looking at the screen from the perspective of the camera’s viewpoint; empirically, then, those views are inferred to represent attention to the content on tire screen. Note that if software can identify the source of the image (whether by identifying the platfOonn in use employing its embedded camera or by identifying tire video source of a connected camera), that can assist programmatic improvement in identification accuracy through limiting the cluster of camera positions to those known to be available to the device.

[0248] It may also be possible to conduct in-session machine learning, assessing the separate video frames or series of frames to identify the total scope of presented eyeball optical vectors over a span of time and more highly scoring earlier eyelines as likely to represent initial attention to tire screen. Diverted eye contact that repeats to fixed locations may be an indicator of multitasking; this can represent a weakness in content, user fatigue or other such factors that, if widespread among other participants, provides presenters or organizations with understanding that can lead to improvements in the overall video-connected communication experience. Using the real-time recognition of direct versus diverted eye contact to score overall interest / engagement levels among the individual participants in virtual meetings.

[0249] Regardless of the methodology through which it’s possible to recognize whether the gaze of a participant in a connected video communications venue (e.g., video chats, video phone conversations or online virtual meetings) is directed at a device’s screen or diverted elsewhere, there is value in assessing it.

[0250] It is natural and normal human behavior to look directly at interesting or otherwise engaging images, so the ability to track overall interest and engagement through an automated assessment of direct versus diverted eye contact between each participating user and that user’s video communication device display screen.

[0251] Even simple implementations of such tracking have value. For example, a presenter seeing on-screen graphics to indicate the immediate level of direct eye contact can use such real-time feedback to adjust and adapt the style or pace or energy of the presentation and be immediately aware of the presence, direction and scope of any change.

[0252] If a system makes log entries (for which circumstances may require anonymization), an individual or organization can examine these to determine at what duration general interest in a typical presentation tends to fade, which is suggestive of time-limiting individual ‘'platform time” as an informal standard.

[0253] Similarly, a graphic display of log data may be synchronized to a video playback of recorded video communications to help identify the interest-resonance of various categories of topical matter across typical participant populations. This can identify’ topics that draw minimal (if any) interest as best addressed through mechanisms other than connected video communications.

[0254] If an organizational trainer is available, a private review of presentation videos with synchronized interest log score displays can help identify techniques that various presenters have chosen when seeing attention scores dip (e.g., changes in the presenter’s style, pace or energy) that have empirically proven effective with a selfsame audience. (Anonymization through generalization and without sharing individual performance examples will be appropriate for many circumstances), but the validity of training observations and recommendations remain valuable.)

[0255] Far beyond any “entertainment” value or “boredom avoidance”, improvements in participant interest and engagement levels offer tactical and strategic benefits to both structured and informal organizations. Higher levels of interest and engagement allow effective communications to take less time and can allow a reduction in the total time spent in (“lost to”) virtual meetings; research has shown that this can also reduce participant fatigue and increase creativity in ideation. Stronger and more consistent engagement in content can also reduce widespread losses in focus by meeting participants who multitask (e.g., texting, writing or reviewing or responding to email, working with online forms or playing games) while connected video communications continue.

[0256] Methods for displaying overall real-time group interest / engagement / attention levels.

[0257] When any methodology is in place that provides data regarding the relative interest / disinterest of members of any group of participating individuals, this level of engagement versus apathy has neither immediate nor eventual value or benefit absent some reasonable way of presenting that infonnation with intuitive clarity. Several methods for doing so follow.

[0258] In a live or online venue that involves a lectern, pulpit or similar facility there are several ways in which lights can provide interest / engagement feedback. The simplest is a single one-color light (red being the most likely color) to alert a presenter to fallen interest / engagement levels; alternately a single one-color light (here, blue or green or white seem best-suited) can indicate that interest / engagement levels are being sustained. A single light capable of depicting multiple colors may appear red when these levels are low and green when high; while additional colors are possible, expect a broad variety to appear less intuitive, ergo more confusing.

[0259] A similar approach may be taken with arrays of multiple single-color lights, e.g., red and green if two, red and yellow and green (especially in a familiar traffic signal motif), or all one color arrayed in a bar-graph or tachometer configuration.

[0260] Wired or wireless connections to body-wom vibrational alerts can alert to waning interest / engagement, and the length or number of vibrations in a pattern can indicate the severity of a loss of interest / engagement / attention. Other telltales are also possible, including meter faces or the motion of an indicator flag.

[0261] The above methods can readily be deployed for live meetings where participants’ faces are all within the view of local cameras. While the same kind of physically present alerts may be practical for presenters during geographically distributed virtual meetings, there are elements of such online meetings that, while making some of the above alerts more complicated or difficult to deploy, also pennit additional alternatives with their own benefits.

[0262] Where physical meetings might display interest / engagement / attention feedback through physical devices, virtual meetings require the presence of an electronic platform (e.g., smart phone, tablet or computer) for their very existence . These devices all provide their own display screens, and thereby, a venue for displaying text or graphics for immediate interest / engagement / attention feedback. Examples of “dashboard” style displays include those that mimic speedometers, thermometers, bar graphs, pressure gauges and so on.

[0263] Examples of face graphics depicting various levels of attention versus inattention include cartoon, photographic or Al images (e.g., rapt, wide-eyed, eyes-front, eyes askew, yawning, sleeping), one or two eyes (e.g., open wide and centered, looking away, droopy lids, lids shit), weather scenes (e.g., frosty and barren, green but raining, sunny and florid) or even single words or short phrases (e.g., all yours, tuning in. losing them, La-La Land).

[0264] In circumstances where the presenter’s display has content not seen by other meeting participants, a graphical icon for attention may, for example, use color, size, animation or other attributes that reflect its own level of perceived interest / engagement / attention (or excitement) to communicate that of participants.

[0265] Using real-time Al to recognize qualitative facial responses among virtual meeting participants to determine the aggregate of attitudes toward ongoing content

[0266] Presenters in online virtual meetings have no easy way to judge in real time how positively or negatively their presentations are regarded by participants in those meetings, a circumstance that the machine learning facet of Al (artificial intelligence) can address.

[0267] Such positive and negative response scores can be accomplished through machine learning upon completion of a preparatory learning cycle, during which the programming is exposed to a very large number of images of images of faces. Pattern recognition can then resolve types of facial expressions reflecting positive, neutral or negative responses, and may include a consideration of “unreadable” in special circumstances, such as when the face is covered or obscured by fabric or a mask. Additional cycles can further resolve this into, ultimately, degrees of positive or negative attitudes observed within individual images.

[0268] The information source that permits this is the meeting software’s presentation of video images of the faces of participants; noting that not all participants allow video displays of their faces, assessments can only address those whose faces are displayed, w hich may or may not be indicative of the attitudes of the entire participant population.

[0269] The sampling of facial images must include a broad spectrum of ages, races, skin tonality, head shape, eye shape, makeup, maxillofacial circumstances, ethnicity and so on. Note that no identification of any such attribute is ever a requirement for the analysis of positive or negative responses; at its simplest, this might be considered the recognition of smiles versus frowns, although the ultimate result is far more nuanced and includes subtler factors. This is not to suggest that even' presentation benefits by always eliciting positive responses. A presenter may wish to motivate participants to act against some perceived threat or hardship, for example, that presents a more relevant focal point if characterized darkly, thereby eliciting strong negative scores. The point of tire subject methodology here is that it provides a mechanism for providing both positive and negative scores in real time, and whichever may align with the intention of the presenter, both provide important feedback regarding success.

[0270] Methods for displaying the aggregate attitude of participants in virtual meetings in real time.

[0271] When methods or mechanisms can be realized to determine the aggregate attitude of participants in virtual meetings in real time, that aggregate attitude is only one aspect of such scoring and considerations of alternatives and options in displaying such information can impact its usefulness.

[0272] For example, it would be counterproductive to present a lengthy all-text summary, because such an approach is likely to interrupt a presenter’s train of thought, if not the presentation itself. By way of example, consider how a presenter might be oflfput by seeing one of these messages:

[0273] • The participants are displaying mixed emotions, with some of them slightly enjoying what you are presenting, some finding it slightly unpleasant and many showing no reaction either way.

[0274] • If you are now attempting to anger your participants with descriptions of very unpleasant circumstances, you are enjoying great success; however, if you are expecting a more positive response, then the opposite is true, and your best immediate option is to feign sudden laryngitis.

[0275] Both are unacceptable for several reasons, not the least of which is the time it takes to read while attempting to present. Also, models of brain mapping show that graphics and text are associated with different hemispheres, and that the introduction of a textual message as above is especially likely to interrupt the presentation of other forms of spoken, verbal messages.

[0276] Eliminating textual messages (even those shorter than the above examples) places a feedback focus on graphical icons, symbology or other representational devices to communicate a real-time status and, as is obvious from the textural example, those that allow understanding at a glance are more desirable than those that need longer cognitive engagement to accomplish interpretation.

[0277] Broadly, graphical representations of interest can follow one of these paths:

[0278] • A single composite indicator: for example, a bar representing the total number of faces being evaluated (less those that are unrecognizable) with some percentage of its length in green representing positive attitudes, some in black or gray for neutral attitudes, some in red for negative attitudes. This approach allows smooth, analog changes to reflect changes on participant attitudes as the presenter alters content or presentation attributes; in contrast, a depiction of a face in which, for example, eye, eyebrow and lip shifts reflect changing attitudes is less smooth, jumping from one look to the next and taking more time for the presenter to recognize and interpret.

[0279] • Two or three separate indicators - if two, showing the relative positive or negative impact - if three, including neutral responses. A more intuitive depiction might involve a 2-dish balance scale with the size of a red item in one dish represent negative readings and changes size as attitudes shift, with a similar response for a green item representing positive in the other dish also shifts - since, with these, the tilt of tire scales also shifts, the presenter’s understanding of response status changes is very close to immediate.

[0280] • Where scoring includes levels of intensity, the shades of color (red or green in the above examples) can become bolder as scores for each increase.

[0281] Presentation of left-right reversed “mirror image” video of a user as captured by a camera behind a transparent display screen showing only external graphics that are right-reading to the user results in a composite image of the user with those graphics appearing right-reading to those elsewhere seeing the composite image, and other relevant attributes.

[0282] When a front-projection screen can be made transparent for a camera behind the screen while still providing a vivid image for a projector in front of the screen, and with appropriate optical measures taken to prevent the projected image from shining through the screen, that projected image can remain unseen by the camera. The camera can see a user standing in front of the screen, and the user can see the image projected onto the screen. For the convenience of the user, that image may be properly viewed as right-reading and not mirrored; left remains on the left and right remains on the right.

[0283] When these two images are fed into either a connected computer or into a studio or control room video mixer / switcher / fader, there is a productive outcome to presenting a mirror image of the user, as fed from the camera behind the screen, as composited (superimposed) on the same right-reading graphic being fed to the projector at the transparent screen. With proper alignment of the camera (ideally even with the centerpoint of the transparent screen) and the projected image, and with normal full-frame alignment of the composite image, the location on the graphic on the transparent screen at which a user points will be the same point at which the mirror image of the user in tire composite image will be seen as pointing on the composited graphic.

[0284] This effect is true both in a virtual meeting context and in a broadcast or recorded video context; however, there may be different requirements in these two circumstance. When the video frame size that the camera behind the screen captures matches the projected image size on the screen, then the camera can capture the user pointing to any and every screen position; when the camera frames a smaller area, pointing is only visible within that area (because the user’s fingertips will be outside the camera’s field of view). As long as the user understands the limits of such imaging constraints, natural gesturing remains easy to execute, especially in tire context of online virtual meetings and gesturing toward elements of a displayed graphic. For applications like broadcast weather where a displayed graphic is, for example, an animated overlay of satellite imaging over a local map, full frame will provide a more satisfying experience for the user, enhanced by the forecaster’s sustained facial presence. (For once, forecasters don’t have to turn their backs to the audience).

[0285] Note that because the image of the graphic is composited with a mirror-flipped image of the user, it is inadvisable to wear clothing that displays lettering or other asymmetrical designs, since those will appear to be backwards as seen by meeting participants or video audiences.

[0286] Automatic video mirror-image presentation of a virtual meeting participant’s face immediately before a virtual meeting or video recording begins.

[0287] Referentially credited to Kim Scott’s Radical Candor (an approach to management that balances "‘challenging directly” with “caring personally”) as inspiration, the “lettuce pact” refers to an informal agreement to privately inform others if some potential embarrassment in their appearance (or presence) emerges. While lettuce in one's teeth is largely a metaphor, in the context of live or recorded video connections with others, presenting such unintentional grooming gaffes can be of consequential benefit to participants. It avoids distractions, diminishment of perceived stature, unspoken belittlement and other negative impacts on the productivity of the otherwise purposeful communication.

[0288] While one traditional approach to prevention involves a detour to a restroom mirror or forming the habit of using a mirror kept in the workplace as a safeguard, the methodology described herein allows technology to take the initiative in a preventive but not public role.

[0289] Where a camera exists immediately adjacent to a display screen or directly behind an otherwise transparent display screen, and where video imaging of a user’s face is scheduled to soon appear, or where a user can take the initiative before such imaging is shared, this camera and screen combination can provide a mirrorlike facility.

[0290] A left / right-reversed “mirror” image of the user can be shown on the device’s display screen.

[0291] Where the relevant event is a scheduled virtual meeting, this facility may also be scheduled to begin at a user-scheduled or default length of time before the meeting’s scheduled beginning or triggered manually by the user. The software that presents this mirror-image mode may also support a user-configured option to emit an audio, visual or vibrational alert to the soon-upcoming beginning of the meeting. Such pre-event alerts may also apply in those broadcast circumstances where the user's presence on-camera abides by a predetermined schedule.

[0292] The metaphor of lettuce on the teeth extends to facial smudges, makeup, eyewear or facial jewelry not intended for showing in the context of a meeting, misadjusted collars or neckwear and other user-undesired attributes of personal appearance.

[0293] Automatic adjustment of one or more attributes of video images when compositing superimposed images for presentation to virtual meeting participants for default or user-specified effects.

[0294] Two or more images combined within a common space within a single visual frame are considered as “composited” or “superimposed” when it is possible to see elements of all images within the same view. The appearance of such images is reminiscent of the effect that might be achieved if they were to be separately printed on panels of transparent or semitransparent films and then stacked.

[0295] It is rare for any two images (even more rare for three or more) to be composed in a way that makes their elements clearly discernible when superimposed. In somewhat technical terms, perceptual “collisions” may result from inconsistencies created by brightness, contrast, pixel density, color saturation, moire, additive color shifts, additive brightness highlighting, occlusion and other factors.

[0296] This set of clarity challenges creates even more difficulties for combining video images than for combining still images; in virtual meetings, for example, video is often presented at 60 frames per second, meaning factors to improve the clarity of overlayed images must be adjusted 60 times per second. Complicating that, they should be performed in such a way that the resultant video of superimposed layers does not display frequent shifts in the net imagery because those shifts might themselves become distracting.

[0297] For especially difficult combinations, software can take advantage of the human eye-brain combination to integrate multiple separate images viewed over time. (This is, in fact, why movies, videos and animation appear to display smooth motion when they are in fact showing sequences of individual images). If software can determine that the true frame rate of two video sources is at least 60 frames per second for both, then it can present these two sources in alternate frames and the eye can readily combine tire separate images. Note that either one or both can be combined with a third, constant overlayed image, but that it is inadvisable to permit displayed frame rates of 20 fps or less since those can trigger medically or behaviorally undesirable side effects in many individuals. In almost all circumstances, the alternate-frame approach can be rendered as unnecessary by appropriate identification of the attributes of the content of individual image frames and the resulting deployment of curative adjustments to them that do not compromise their ability to impart meaning.

[0298] Specific to online meetings, where there are manifest benefits to maintaining constant eye contact, it is reasonable to presume that one of the two images being combined is video of the face of the user. There is a loosely defined “T” shape to the region of interest in depicting a face that involves the eyes, nose and mouth, and software can prioritize representing the facial image in that region while permitting more flexibility in compromising clearness when displaying outer extremes of the face or torso or the captured background behind the user.

[0299] Where software algorithms can determine the best outcome of combinations of overlayed pixels from two separate images fonning a third image, and because the image of a user’s face does not undergo frequent or abrupt changes, and because presented overlay content is unlikely to present frequent or abrupt changes, the practical demands of processing multiple images for virtual meetings are less daunting than they appear. Even with two 4K UHD images, each representing 8.3 million pixels 60 times per second (which mathematically suggest approximately 500 million pixel-manipulations per second), these circumstances let software follow simpler processing paths. When average attributes of any next frame are like the current frame, processing requirements are zero to minimal, and the software can preview many future frames before they are displayed.

[0300] Under ideal circumstances, video images selected for compositing can be audited by their own authoring software to suggest edits, choices or adjustments that mitigate against major changes being required during the virtual meeting in which they will be presented.

[0301] Separation of video communicated to virtual meeting participants from video displayed to the local user through software.

[0302] Contemporary virtual meeting software applications or utilities follow no ubiquitous standards in terms of the number or the contents of windows on a user’s screen(s). Uris software or adjunctive utilities may make modifications to the video image of the user that is shared with other meeting participants, but seldom does anything other than introducing or altering the image of whatever background is behind the user. A user’s screens may show a collection of images of other meeting participants, an image of the user as presented to the meeting and an image focusing on the person currently presenting or speaking. There are many benefits to allowing purpose-designed software running on a user’s own local computer to separately manage the image going to the meeting from tire image presented to the user in a new window (and ideally, on a separate screen).

[0303] For example, the added local screen can, without sharing any of these elements with others in the meeting, selectively display (through software configuration selections and software-managed content storage) count-up and / or count-down timers when presenting, along with theme notes, a scrolling prompter script and thumbnail images of prepared presentation modules. It may also offer (if such facilities are in some way available to the software) real-time scores on meeting participant attention and engagement, and on their positive or negative regard for the content being presented.

[0304] At tire same time, the video of the user that feeds through the meeting software as if viewed on tire user’s webcam can include superimposed, composited or otherwise layered content.

[0305] The single predicate to facilitating such improvements to the virtual meeting experience fundamentally depends on the ability to separately manage and assemble a user's meeting-related video as seen by the meeting from the same user’s meeting-related video as seen on his own separate display.

[0306] Compositing through software the superimposition over other images of a presentation count-up timer and / or a count-down timer and / or a combination count-up / count-down timer so as not to be visible to other virtual meeting participants while fully visible on video displayed to the presenter.

[0307] When presenting to a virtual meeting, or when viewing a recorded video presented by another participant in a virtual meeting, it can be useful to know ho ' much time has elapsed since beginning. When presenting, tracking elapsed time through the addition of elapsed time infonnation to a display visible to the presenter can help achieve such common goals as enhanced communications efficiency, providing relevant information in a less-than -usual length of time.

[0308] Further, any mechanism that can make this visible only to the presenter can provide that benefit without simultaneously introducing the infomiation as a potential distraction to other meeting participants, which could have the effect of reducing their attention to and thereby absorption of important infonnation within the presenter’s content.

[0309] The specific format for displaying this timing readout is not generally relevant, providing it can be instantly and intuitively recognized and understood by the presenter. It might (for example) appear as a digital numeric display, an analog clock face, sand in an egg timer, a burning fuse, a character in distress or in any other symbolic, iconic or metaphoric representation. Optimizing its visibility to the presenter while minimizing the potential for distraction when the presenter observes its immediate reading strongly suggests incorporating it into the image on the presenter’s screen(s) that is most relevant to tire presentation of content to participants in the meeting. This suggests compositing the time-related display into whatever images are present on that display. (We may presume that the presenter is already familiar with the information being presented and therefore any accidental masking or occlusion of the presented image is not counterproductive). The standard methods for doing so include the introduction of a chyron (electronic caption or “crawl”) band or stripe into the top or bottom of the presenter’s displayed image, window or screen, or image superimposition (optionally involving some level of transparency for the timing display), or superimposition with no partial transparency (keying or inserting).

[0310] Awareness of elapsed time may also help improve presentation skills if the presenter uses this information to set goals for greater communications efficiency within progressively shorter-duration presentations.

[0311] Similarly, where there is a desire to fit a presentation within some specific length of time, a count-down timer that is visible only to the presenter may also contribute to a presentation achieving its desired communications efficiency without distracting other meeting participants. The motifs for presentation already described also apply to a count-down time display.

[0312] Also, while it's possible to separately include both count-up and count-down timers as separate displays, the goal of immediate (“at a glance”) intuitive recognition is better served if the readings can in some way be joined or integrated, as with, for example, a progress bar display, showing the total intended meeting duration with a visible moving status marker between the ends and optionally numeric readings of the time elapsed and time remaining.

[0313] Compositing through software the superimposition over other images of a current virtual meetingduration count-up timer and / or a count-down timer and / or a combination count-up / count-down timer so as not to be visible to other virtual meeting participants while fully visible on video displayed to a local user.

[0314] Virtual meetings include participants who are not all present within a single physical location but who are all willing to jointly participate during a stated length of time, expressed either as a starting and ending time (with all times adjusted for the participant’s time zone for synchronous participation) or as a starting time and intended duration. Either method makes these three factors (starting time, intended ending time and intended duration) known. For each participant in such virtual meetings, it can be useful to know how much time has elapsed since beginning. It can also be usefill to know how much of the allocated meeting time remains. Also, for such factors to allow a focus on the content of the meeting without discernable distractions for common clockwatching activities (e.g.. turning to see a clock, diverting attention to locate and read a computer toolbar’s time display, looking at a wristwatch or checking the time on a cell phone), this temporal information is best integrated into a display window that the meeting participant is already viewing. A subtle, nondistracting time display can help aid the focus and engagement with content of every participant with a positive work ethic.

[0315] Including numeric, textual or graphical depictions of elapsed and remaining time on an already meeting- involved display window can be accomplished through software compositing the time display atop the content already being sent to that display window. The standard methods for doing so include the introduction of a chyron (electronic caption or “crawl”) band or stripe into the top or bottom of the presenter’s displayed image, window or screen, or image superimposition (optionally involving some level of transparency for the timing display), or superimposition with no partial transparency (keying or inserting).

[0316] The specific format for displaying this timing readout is not generally relevant, providing it can be instantly and intuitively recognized and understood by the presenter. It might (for example) appear as a digital numeric display, an analog clock face, sand in an egg timer, a burning fuse, a character in distress or in any other symbolic, iconic or metaphoric representation.

[0317] Also, while it’s possible to separately include both count-up and count-down timers as separate displays, the goal of immediate (“at a glance”) intuitive recognition is better served if the readings can in some way be joined or integrated, as with, for example, a progress bar display, showing the total intended meeting duration with a visible moving status marker between the ends and optionally numeric readings of the time elapsed and time remaining.

[0318] Collecting, organizing and displaying a user’s next scheduled virtual meetings with date, time and choice of virtual meeting software to use, and optionally including a way to display next-meeting listings immediately following the end of a current meeting as a composited superimposition over a local display window that is fully visible to the user but not shown to other meeting participants.

[0319] Virtual meetings are today an established standard within and beyond business practices, but beyond empirically rare, entirely coincidence-driven circumstances, no single, reliable mechanism exists to automatically cite both the when (date and time) and the where (which meeting software to use) of every next meeting. The methodologies presented here address this shortcoming. One occasional but widespread reason that invited and confirmed participants may fail to join a virtual meeting is that it does not appear on the same scheduling or calendar application, utility or facility that they consult most often. In turn, one contributing reason for this may be that the meeting is scheduled to connect through an online meeting application, utility or facility other than the one through which most of that individual’s meeting bookings occur. For example, where most of an individual’s meetings connect through Teams, a Zoom meeting my be forgotten or go unnoticed (or vice-versa). Also, for example, an individual or organization may choose to standardize on a scheduler or calendar that does not enjoy existing links to meeting software schedules (e.g., a meeting scheduled through Teams intrinsically appears on the same user’s Outlook calendar in a Microsoft environment while meetings scheduled to occur through Meet, Zoom, Webex or others tend not to be likewise automatically posted).

[0320] In addition to multiple online venues whereby virtual meeting connections form, the meeting’s associated organization and its protocols may be relevant to its own employees, but it’s just as likely that their functional roles involve interacting with (if not being) participants from other organizations. Across all participants of most meetings, those separate roles create a chaotic canvas beyond company personnel that includes, for example, third-party consultants for whom the organization is a client, and suppliers or customers or other variations of buyers meeting sellers. When viewed from the context of the meeting, designing or accomplishing a homogenized approach across all participants becomes an unlikely to impossible task. It is the de facto nature of virtual meetings that denies a de juris (mandated by rule) solution. In short, there is no shortcut to accurately telling even' participant what meeting (and where and when) is next on his or her agenda.

[0321] The methodologies that do permit this exist not in the context of meetings but in the context of individuals. Each such methodology involves finding one or more ways to monitor the scheduled meetings across all virtual meeting applications, utilities or facilities available to each user. The same is true across all scheduling or calendar utilities, facilities or applications. A variety of established information harvesting techniques (e.g.. using published APIs for such applications, or bringing up and reading this information when exercising them through automation) may be included in such data-gathering operations.

[0322] The data thus collected can reside in local and / or Cloud-based upcoming-meeting logs, organized as a database. Automatic monitoring can detect conflicts and alert the user to resolve them; the chosen meeting is included, others flagged as conflicts, and ‘‘regrets” messages automatically sent to tire source of invitations to those. Hie software managing meeting listings can offer its own user interface; its user can also elect among installed calendar / schedule applications for automatic updating from this data collection and management processing. The creation through software of machine-storable single-screen prepared presentations able to accommodate combinations of static and dynamic content available to that computer.

[0323] Before the early twenty-first century pandemic, most computer-involved presentations used a metaphorical slide show' venue (e.g., like PowerPoint) or video; these may involve non-static (moving or changing) content or true static content (e.g., depictions of pages or still images) or a hybrid (e.g., still images viewed with imbued motion through camera or object movement). During and since the pandemic, with the rapid escalation of usage of virtual meeting mechanisms (e.g.. Teams, Webex, Zoom, etc.), their primary dependence on so-called “screen sharing” techniques allowed the inclusion of computer-based application window's. This w'as often helpful and useful, but often illegible to remote meeting participants (often because of scale, for example, with typography too small to read in a remote window) and difficult or impossible to simultaneously display along with relevant secondary or tertiary contributing information that could have aided understanding.

[0324] Since the beginning of windows-based computers, it has been possible to expand a window to the full size of a display screen. This describes the complement to that approach, wherein software can integrate content that w'ould normally occupy single or multiple windows into a single window with the look and feel of a screen-full (note: not the same as full-screen) display.

[0325] When software can let a user create (for the user’s own presentation, for presentation by others or for publication) a single entity incorporating the content of one or more windows into a single window showing everything an entire computer display might present, the new unitized entity’ offers several benefits and advantages.

[0326] Because these authored screens retain the format of a full computer display screen, and because the gathering of content for any one of them is reminiscent of selecting the elements that manifest a meal in a cafeteria, yve have been using “tray” as an internal term yvhen referring to each such presentation screen. As a convenience, and to avoid tire confusion among various companion terms (e.g., yvindoyv, display, screen, presentation, prepared content), we will use the term “tray” here. It is not in tire scope of this description to discuss the authoring of such displays.

[0327] A tray could simply (like screen-sharing) incorporate a single yvindow. Even in that regard, a tray may often be easier to locate for reuse, to include in a “collection” (library) or to be published (e.g., as part of a pharmaceutical company’s briefings to prescribers about a neyv medication, yvhere such a briefing may originate from multiple elements of or associates of that company, or yvhen separately addressing different groups of participants across multiple meetings). The ability to represent any combination of still images (also including documents or titling) with moving images (e.g., recorded video, linked live video, PowerPoint-like transitions, hybrid animation, etc.) affords an enriched ability to combine images, text and more and, through these multiple methods of parallel communication, improve the personal engagement of virtual meeting participants.

[0328] For any one user, each such tray is likely to embrace the same aspect ratio and resolution. When the full available resolution of the user's display is the default for the resolution of the tray, virtual meeting participants benefit from contemporary computer graphics handling that optimizes the display of higher resolutions for best viewing in smaller windows. Among all the attributes of any single tray, anticipate those of highest interest to include the tray name, subject, creation date, file size and playout duration.

[0329] This unitized presentation format can apply both to presentations within virtual meetings and to presentations made in person to one or more people. It can be authored to require no introduction (for example, when presented in support of an “umbrella” topic) or to stand alone. It should also be possible to export into formats that can be played from web pages or that can be downloaded and viewed later. Its directly-authored format may be among existing standard computer file formats (e.g., MP4) or may be a new purpose-built formula (in which case a file player application, utility or facility could be made available for relevant operating systems).

[0330] The creation through software of a library system for organizing and maintaining multiple machine-stored single-screen prepared presentations available to and callable from a user’s computer.

[0331] In earlier computer eras, the term “prepared presentations” most often applied to either videos or slideshow applications (e.g., PowerPoint). Since the Pandemic, as the most-often-used presentation venue became virtual meetings, fewer presentations are prepared, more are done ad hoc through the meeting software facility popularly known as “screen sharing”. Screen sharing could involve unpredictable sources (e.g., web pages. PDF pages, videos, Word documents, spreadsheets or CAD). While all such content might ultimately prove reusable, it is often difficult to locate anew, and that difficulty is exacerbated with the passage of time.

[0332] When software can let a user create (for the user’s own presentation, for presentation by others or for publication) a single entity incorporating the content of one or more windows into a single window showing every thing an entire computer display might present, the new unitized entity offers several benefits and advantages. These include tire ability to electronically store and recall it and the ability to convert or export it into file formats that allow online web display or download-based publication. It is not in the scope of this description to discuss the authoring of such displays.

[0333] Because these authored screens retain the format of a full computer display screen, and because the gathering of content for any one of them is reminiscent of selecting the elements that manifest a meal in a cafeteria, we have been using “tray” as an internal term when referring to each such presentation screen. As a convenience, and to avoid tire confusion among various companion terms (e.g., window, display, screen, presentation, prepared content), we will use the term “tray” here.

[0334] Among all the attributes of any single “tray” file, anticipate those of highest interest to include the trayname, subject, creation date, file size, playout duration and work status (e.g., draft, in-process or final). Tray authorship is most likely to reflect the highest screen and resource resolution available to the user authoring that tray. Ideally, the tray authoring software will create a graphical icon representing the first image that the tray presents when played. Ideally, the tray authoring software will by default save its work in a single folder, or within subfolders of that folder; the folder or folders may be local, online or both. Software facilities within or connected to the authoring software can build, maintain and manage the connected file library- or libraries. Similar software facilities can manage the extent to which files are shareable with others. Companion software facilities for those who are not using such tray' authoring software can let them selectively connect to, search and use content from available libraries when either preparing their own presentations, viewing content (because of, for example, having missed a meeting), or importing content for other purposes (e.g., online or optical disc publication).

[0335] An authoring and editing system - optionally involving templates, versioning and assistive guidance - for conveying the appearance of single or multiple windows within single, unitized prepared presentation modules.

[0336] On any given personal computer monitor or display? screen, most users enjoy the ability? to display a single window or an array of multiple windows. Those windows remain independent and may be independently moved, resized, closed, appended, content-zoomed or moved to the foreground or background in that display. Tire challenge of authoring a single element containing a unitized collection of window-like elements is justified by the applicability of such unitized frameworks to many purposes: these include reusable presentations for sharing within virtual meetings, web-viewable or web-downloadable presentations, exportable presentations for sharing over public (e.g., YouTube) sites and more.

[0337] Because these authored screens retain the format of a full computer display screen, and because the gathering of content for any one of them is reminiscent of selecting the elements that manifest a meal in a cafeteria, we have been using “tray” as an internal term when referring to each such presentation screen. As a convenience, and to avoid the confusion among various companion terms (e.g., window, display, screen, presentation, prepared content), we will use the term “tray” here.

[0338] Software to author such unitized “trays” begins with the ability to include content from a variety of sources and in a variety of ways. Full or partial pages from sources like documents or spreadsheets or websites, for example, might be directly read from their native files, or read through conversion to an intermediate format (e.g., PDF or image files like JPEG), or captured as graphics through mechanisms like screen clipping or screen saving (including screen-saving browser pages). Most of these represent static content, but similar approaches can also bring in dynamic content (e.g., video files or CAD rendering animation). Some content (e.g., PowerPoint presentations) may include automatic or triggered “slide” (page) transitions as well as elements within pages that incorporate dynamic highlighting (e.g., through elements like highlights, splashes or shifts in size or color).

[0339] Among any collection of candidate elements, good design practices suggest choosing a main purposeful element and supporting it with ancillary content (e.g., illustrations, brief expert snippets, images of predecessors it will replace). For users with a strong understanding of (“eye” for) design, authoring can be a freehand exercise; such a user can have full control over the sizes, positions and other attributes of individual elements as well as shifts in their layout. Those and other users will find it helpful for authoring software to present a selection of templates to offer a head start in organizing the unitized collection

[0340] With both static and dynamic (or quasi-dynamic) content potentially included in such a unitized presentation, an authoring system must also incorporate an ability to deal with trigger and shuttle controls; trigger controls, for example, let the presenter initiate shifts in individual or multiple element attributes (e.g., size, layout, appearance, disappearance, positional swaps, transitional motion, fading or the start of a video playback or the transition to a next PowerPoint slide); shuttle controls can support forward or reverse transitions through events (e.g., going to specific elements of a presentation in response to a question, or to a specific playback point within any dynamic content element).

[0341] Such authoring software can also coordinate with internal or external scrolling prompter text displays to help a presenter remain seamlessly synchronized between the displayed image and that user's narration.

[0342] Editing is also an important function of most authoring systems. Any individual element within a unitized collection may benefit from alteration if modifying it can, for example, remove distractions, increase legibility or accomplish its purpose with more clarity. As seen across almost every form of authoring, while an initial first pass may accomplish an acceptable result, such a “first draft” tends to offer room for improvement. Most of tire same composition adjustments that go into authoring must remain available as editing tools. During both authoring and editing, software should support versioning, meaning file saves of work in progress such that a user can quickly revert to an earlier version when finding it superior to attempted changes since.

[0343] Authoring and editing software can also optionally provide assistive guidance if it can determine (e.g., algorithmically or through a deployment of artificial intelligence) that the user’s work product can benefit from a choice other than one the user makes. For example, reframing or masking an element can eliminate irrelevant or illegible content within it, repositioning elements may better support the graphical-arts concepts of eye-fall and eye-flow, brightness or contrast adjustments may improve clarity, and so on.

[0344] Automatic recognition and identification of circumstances where full- or partial-screen content within single-screen presentations will result in visual elements being too small to be easily perceived when presented on a secondary display.

[0345] Captured images of documents, screen content, charts or other computer-accessible content as well as original or copied images created as images are ubiquitously assigned to many duties by computer users. While most such images were well-crafted for their original purpose, their new purpose may introduce constraints that compromise their legibility in these secondary roles. One frequent example of this occurs when what was, in the original document, small but legible typography that loses its native legibility when resized or rescaled to the new purpose; consider, for example, trying to show the entire text of a newspaper page on a postage stamp; while the example may seem to exemplify reductio ad absurdum (in logic, a reduction to absurdity) it is more a reduction to absurdity to expect the result to remain legible. Software- fostered recognition of circumstances can reduce the legibility of typography and other image elements to absurdity.

[0346] The need for this in typical computer usage has been marginal because across the population of all computer end users, most content authorship occurs within application frameworks where size defaults exist that aid compatible legibility between creator and receiver. Normal usage of mainstream applications (e.g., as word processors, database managers, email or texting facilities, spreadsheets and so on) may still present occasional legibility issues but overall do not.

[0347] Scaling issues became more commonplace with the growth of websites, and much more commonplace with the growth in virtual meetings, especially during screen-sharing. In sharing either their entire screen or a single window, a virtual meeting participant may enjoy a more legible local view of the shared content than would be visible to those whose meeting softw are is either not displaying full-screen or is displaying on a smaller screen; it is most vulnerable to elements becoming illegible for participants on smaller smart phones. Both the participant who shares a screen and the participants seeing it may have display zoom options available, but the need for any such adjustment is often distracting and may interfere with effective communications between the user presenting and the other participants trying to understand the meaning of the shared content.

[0348] When screen sharing is performed ad hoc, absent time to prepare the source content for clarity, elementary use of core OCR principles can often detect tire existence of alphanumeric characters within an image when those are more than 5 pixels high. Software can suggest immediate cropping or zooming; this must then be the presenting user’s choice because human decision-making can recognize whether either can be done without eliminating the tactical content focal point of sharing the screen.

[0349] When content can undergo authorship or editing before being shared, software can accomplish more analysis and guidance. It can identify subsections of images that might, individually or in combination, be better candidates for representing the desired point(s) of information; once that is confinncd for identified image-parts, software can suggest resizing and repositioning these into image combinations from among which a user can select a favorite.

[0350] Advanced character recognition is not a requirement for this approach to result in improved local and remote legibility of shared screen content, and the same principles can also extend to content being considered for other applications (e.g., document or web page design).

[0351] Automatic preparation of thumbnail views of unitized or single-window prepared presentations.

[0352] Computer-based presentations that arc saved as named files or saved within named subfolders arc often identified only by those simple, textual file or folder names. Locating specific presentations is often challenging with only a textual identifier. The more presentations that a user or organization stores together, the more complex the task of locating any one that is newly intended for presenting again or other sharing.

[0353] Complicating this is the breadth of applications that may have a role in authoring, editing, rendering, encoding or transcoding video files, CAD images or animations, photographs or graphics files, as well as the scope of content created by popular office applications, and so on.

[0354] Most computer-view ed presentations are seen in an aspect ratio that is not identical to the defaults for the applications that create them; common screen formats are landscape (horizontally favored) in a 16:9 aspect ratio, with other aspect ratios and many portrait (vertically favored) display orientations also often present.

[0355] Considering the filmmaking tenn, “opening shot”, describing the very first image (not counting credits) seen by an audience. This concept is already deployed for the icons of video files. Extending this concept to single-window prepared presentations is more challenging, given that the applications associated with most such window content already have ingrained protocols for creating their file icons.

[0356] Tire concept, then, of attaching thumbnails to file icons for myriad application type and file locations offers an unproductive approach for, potentially, a highly productive purpose. There is a more appropriate approach in recognizing the likely presence of software applications, utilities and facilities to independently track and perhaps simplify discovery of and access to prepared presentations.

[0357] Our topic here addresses the automatic preparation of thumbnail views of presentation content as an extension of authoring or editing such content, or independently through the services of separate, purpose- built software.

[0358] Software mechanisms to automatically capture a thumbnail from tire first image that a presentation displays include, for example, automated screen clipping during manual or automated playback, or automated capture in concert with the authoring or editing software used to create the presentation.

[0359] Remotely controllable methods for selecting among prepared single-window, single-screen or otherwise unitized presentations for viewing beyond the user’s current computer display.

[0360] When presentations are prepared in advance for local or remote viewing, but not necessarily viewed on the computer window or display from which they are selected, their benefit can be made available through such venues as projectors, browsers, virtual meeting software, connected computers and so on.

[0361] Local issues or concerns (involving, e.g., management, authority, security, convenience or other factors) may lead to policies that result in a preferred computer making selections among multiple available prepared presentations for viewing, sharing or forwarding in any such regard. Physical, ergonomic, convenience, access-control or other factors may place such a preferred computer out of easy reach of the user who makes selections among prepared presentation content choices. It is also possible that the selection-controlling computer will manifest a deliberate absence of its own local user direct interface. In circumstances like these, the use of a remote control offers a mechanism for satisfying the challenges of these combined requirements.

[0362] Noting that organizations in which such stored-content selections are present may need to have such selections made for multiple endpoints from multiple selection-controlling computers, it is important to avoid a ’‘party line” or “crosstalk” effect, where there is no implicit one-on-one relationship (“pairing”) between individual remote controls and individual content-selection computers. One example of accomplishing this without requiring overly specialized signaling devices is through the use of Bluetooth interfaces: where explicit one-on-one pairing cannot be as easily accomplished (for example, in facilities that forbid radio-frequency signaling), it can also be done, for example, through techniques that impose unique command-recognition headers on communicated signals (that might, e.g., be transmitted at selected wavelengths of light).

[0363] If security is of high concern, a remote can also work with the selection-controlling computer to provide unique identification passcodes as needed (e.g., once per day, once per shift or once per project).

[0364] While the selection-controlling computer may be absent any keyboard or pointing device, it can (and should) be able to display choices that can be individually highlighted through operation of a remote control. In circumstances where it is important to not locally display any of the actual selection content as images or refer to it in titling, assigned access codes can be built through interaction between tire remote as a pointing device and an on-screen keypad. When such security concerns do not exist, a small number of onscreen content thumbnails (ideally, with titling) can, using the remote as a pointing device, select an element or a sequence of elements for inclusion.

[0365] Depending on the number of entries and complexity of their categorization, it may be possible to simplify navigation through preselection mechanisms (e.g., hierarchical category menus, often-requested presentation groups or clusters and so on).

[0366] Using Al machine learning and other measures to reduce insertion errors like “scalp sizzle” when synthesizing background images for on-camera virtual meeting participants.

[0367] False backgrounds for such video images as those that webcams provide are favored by many computer users as a way of disguising their surroundings: for example, making a space within a home appear more businesslike.

[0368] One common result of attempts to create such video insertions over synthetic backgrounds, nicknamed “scalp sizzle”, tends to be visible near the top of the head, where small blocks of pixels spontaneously shift between showing the user and showing the background. This is a result of uncertainty by the relatively simple software algorithms that determine where transitions between the live video image foreground and background are positioned, with a tendency to make such determinations imprecisely. (Note: the “scalp sizzle” epithet mostly references areas near the head, but the effect can also occur elsewhere within the image).

[0369] The imprecise assignment of foreground versus background may be exacerbated by poor lighting, poor camera resolution, poor camera sensitivity, vagaries in determination algorithms or a general lack of intravisual cues to help identify’ foreground versus background transition point more rigorously. Within the broader category of Al (artificial intelligence), the methodologies of machine learning can result in improved recognition accuracy. Such methodologies involve allowing computer analysis of very large numbers of images in a way that invites escalating improvements in recognition accuracy. These result in improved recognition algorithms that can then be deployed in improved background substitution programming.

[0370] While machine learning can bring strong improvement, other measures can also contribute strongly to improved results. These include using better cameras with improved attributes (e.g.. resolution, sensitivity, color depth, etc.) and better lighting that, by its nature, helps visually separate foreground objects from backgrounds.

[0371] Using Al machine learning to provide realistic moving images of clothing items as motion-tracked synthesized visual replacements for the actual clothing of a virtual meeting participant.

[0372] Simple existing software operating with video images of people can (albeit imperfectly) detect differences between a person as a foreground object and objects in the background. Such software may understand the foreground object as an outline, within which video continues to pass while outside of which other images (e.g., substitute backgrounds) are added before the combined video is shared (e.g., transmitted or recorded). This is prior art.

[0373] Imperfect analysis leads to pixilation in determining such outlines; this can lead to such artifacts as “scalp sizzle" (where image elements near tire top of the head shift between being considered as foreground and background). Techniques from the machine learning disciplines within the broader scope of artificial intelligence can lead to algorithms that significantly reduce or eliminate such artifacts.

[0374] Similar forms of edge and feature detection have also been deployed for various purposes, from novelty (e.g., substituting an animal or cartoon face) to trickery (e g., deep fakes). This, too, is prior art.

[0375] Machine learning can enhance such edge detection around and within (between sub-elements) of a video image to also track natural motion. For example, the left arm, right arm and torso of a video foreground person may be identically tracked. When so tracked, substitute images may be synthesized by software to substitute for the original image (in our example, of left arm, right ami and torso). The recognition of motion would allow such substitutions to include the natural flow of clothing fabrics.

[0376] In an era of escalated daily involvement in virtual meetings, some users who are expected to participate may be embarrassed by what they happen to be wearing, and such people may elect to keep their cameras turned off, so their presence is indicated only by, for example, an initial or two. With the ability to substitute a chosen synthesized wardrobe, such hesitation may be overcome, with the potential for resultant benefits to the user, to other virtual meeting participants and to the company or companies behind such meetings.

[0377] In the vernacular, somebody who needs to join a meeting when wearing only “sweats” or a “wife-beater” could be seen as wearing an Armani or Brooks Brothers suit (assuming the existing of licensing of such designs), or a tux, or slacks and a sports coat - and the synthesized fabrics would flow naturally as they moved.

[0378] Using transparent, embedded or otherwise unseen cameras within or in front of a display screen so that the act of viewing the screen inherently accomplishes direct eye contact.

[0379] The desirability of full-time eye contact during virtual meetings, video recordings and similar real-time or recorded video activities fostered many efforts to place a user-facing camera at the center of a display. Users who are not trained or practiced at fixing their gaze at the camera instead of at the display will appear within the video feed to be looking at something off-camera. Such apparent diverted attention is often interpreted as inattention or as evasiveness.

[0380] Most video-enabled devices today embed a camera within a display’s bezel or place a camera atop a display bezel. Some smart phones achieve a similar end by placing a camera within the display area but to one side. These techniques continue to result in the exhibition of a diverted gaze.

[0381] Several attempts to center a camera behind a screen were based on transparent LCD or OLED displays; the polarizers native to LCDs significantly dimmed such images; for all such displays, transparency was approached by shifting the active pixel (picture elements) sections to one side while allowing partial transparency from the resulting tiny gaps between pixels. (Cameras behind those screens, even with images enhanced through hardware and software, could not overcome the intrusive “speckles” of the active sections within each pixel). Attempts to mount a camera directly behind a display with a hole drilled through the display to clarify the camera’s image did work for the camera but were more than normally distracting to the user both because of the hole in the image and because of the artifacts (e.g., a ring of light at the edge of the hole from ambient light on the screen internally reflecting to that edge).

[0382] One method that does work involves front projection onto specially micromachined glass; while successful, it is also expensive.

[0383] This describes a strategic approach of embedding a camera within a display, including as examples two tactical approaches to manifesting this.

[0384] INTRA-DISPLAY CAMERA: When display pixels can be minimally separated while separated to the extend that single camera pixel sensors can fit the resultant gap, and front optics can be incorporated into the front of the screen to help coordinate the respective fields of view of these individual camera element, and these camera pixel connections then route to a connected processor, algorithmic handling of the individual pixel images (especially if enhanced and improved using artificial intelligence machine learning techniques), the collection of such spread video pixels around the center-point of a display can result in recompositing an active video image that does not include display pixels within its field of view. This accomplishes the goal of providing direct eye contact for users who simply look at the display screen.

[0385] FORESCREEN CAMERA: When the electronics and substrate and connections of a camera can result in a video imager that is essentially transparent, such a camera can be mounted on the front of a display screen, at its center, without necessarily blocking a user’s view of the display. Display electronics can also adjust the attributes of display pixels directly behind such a transparent camera so that the net image as perceived by the user is only minimally or negligibly obscured by the camera. Uris accomplishes the goal of providing direct eye contact for users who simply look at tire display screen.

[0386] Applications for such transparent or otherwise invisible or unperceived embedded cameras can also extend to applications where the display may not be electronic, such as vehicle windshields, window-embedded security cameras, enhanced gunsights and purpose-enhanced mirrors.

[0387] Selective automatic forwarding by email of content viewed during online meeting, optional logging to CRM.

[0388] Virtual meetings, while frequent and commonplace in business and elsewhere, are nevertheless conducted through software that offers few features that can enhance the productivity of users, advance the missions of the business or escalate the depth or perseverance of the information that it was the meeting’s intent to share.

[0389] It is also noteworthy that virtual meetings are not yet routinely deployed as venues for pre-sale support, technical support or customer support. With a few limited exceptions (e.g., telemedicine), information may be shown (be tools like, e.g., screen sharing) but does not thereby become available to meeting participants.

[0390] Absent any significant “take home” (recoverable) mechanism for such content (and, ironically, given how many meeting participants are identified only by initials and not with live video), the net communication value of many modem virtual meetings is no better than earlier generation meetings over land-line telephone conference calls. With faces showing but without additional content, virtual meetings fit a broadly discommended motif (known to film and video producers as “talking heads”).

[0391] Virtual meetings (at the time of this writing) gain content primarily through the practice of “screen sharing” where anyone presenting to the meeting can choose the content of an open display window and essentially publish it to the meeting. It is reasonable to assume that this may evolve to allow the sharing of unitized multi -element prepared briefings to others in the meeting. This facility is equally applicable in one-to-many or to one-to-one virtual meetings.

[0392] A one-to-one context (to simplify understanding of core concepts) provides opportunities for a dynamically varying sequence of multiple content presentations in response to either party’s expressed questions, interests or other factors; this allows a more fluid, interactive and engaging overall session.

[0393] NOTE: Practical purposes for one-on-one virtual meetings involving one or multiple content presentations include, for example, pre-sales support, technical support, customer support, press interviews, new product introductions to prospective customers or referrers, presentations to prospective investors, hotel virtual concierge stations, telemedicine, direct sales (e.g., sales to homeowner of insulated windows, roofing or blow -in insulation - sales of vacations or time-shares - selections of a Medicare Advantage program - life insurance sales, senior care facility presentations, club membership sales - or requests for donations by charities or political interests - new or used vehicle sales) and more.

[0394] Because online meetings are scheduled events between parties, and because meeting notifications (“invitations”) tend to capture a participant’s e-mail address, that information is available to software either integrated into or running concurrently with the software through which the meeting occurs. The same software can also capture the first or full name of the person making the presentation(s). Such software can also gather codes that identify the sequence of presentations shared during the session; simple tagging can identify whether all or only some subset should be referenced in follow-up communications between parties. It is also possible through any of various mechanisms (e.g., if the request for the meeting is initiated at a website and a form there asks for additional information of interest) to capture product or category interest, location, company affiliation and so on, and all such data becomes available to the referenced purpose-designed software. Where there are thumbnail images available for each prepared content presentation, the software can also recognize and access those.

[0395] When the meeting ends, this software can then automatically initiate an email message from the presenter to the connected party. Here is an example (the example presumes that the selfsame presentation content is addressable somewhere on the organization s ’ website) of what its content might be.

[0396] Thank you for reaching out to us and for taking the time with Jerry (name of presenter) today to leam more about Translators [product or category name). If you’d like to review what was said or share it with others, here are the presentations you saw including links to them on our website, where you can view them again, download them or share those links with others. If you'd like to schedule another virtual meeting, click here. Or you can send me an email or phone me to take our next steps.

[0397] Here's what Jerry shared with you:

[0398] • [Thumbnail 1]. PRESENTATION TITLE. Note on its content. Click here.

[0399] • [Thumbnail 2], PRESENTATION TITLE. Note on its content. Click here.

[0400] • [Thumbnail 3], PRESENTATION TITLE. Note on its content. Click here.

[0401] • [Thumbnail 4], PRESENTATION TITLE. Note on its content. Click here.

[0402] •

[0403] • [Thumbnail ##?]. PRESENTATION TITLE. Note on its content. Click here.

[0404] Thanks!

[0405] Signature

[0406] Firstname Lastname, Title

[0407] Email

[0408] Phone

[0409] Skype

[0410] Organization

[0411] Website

[0412] Location

[0413] Slogan

[0414] The collection of information (potentially including other collected information, like dates and times, originating IP address, etc.) about the meeting requester is also available for integration with an organization’s existing CRM (Customer Relationship Manager) or similar software for a variety of purposes (e.g., follow-up cues, outreach efficiency scoring, conversion to purchase scoring, etc.).

[0415] Such programs are also applicable where the presenter initiates meetings; for example, a pharmaceutical company introducing a new prescriptive may want to initiate outreach to relevant specialists. Also, as stated, they also apply to one-on-many meetings; for example, where a small college or prep school holds virtual class reunions which present either changes to the school or where solicitations for donations may be appropriate.

[0416] Determining through Al or other software mechanisms the number of virtual meeting participant endpoints providing face video within the total number of currently connected participant endpoints and / or among the subset with faces showing, the number showing approval, disapproval, neutrality or disinterest.

[0417] Virtual meetings, by definition, involve the participation of at least tw o parties. There may be a difference betw een tire number of people in a meeting and the number of connected endpoints for the meeting where two or more people share an endpoint (e.g., where one endpoint is a conference room populated by multiple participants, or where two or more colleagues gather in an office to participate over a single device). For that purpose, we must specify the “universe7’ of a meeting in terms of participant endpoints and acknowledge that the total population of people may exceed the number of participant endpoints.

[0418] The association of a face with an endpoint involves either cooperative (e.g., internal integration or API access) or derivative connections (e.g., recognition of circumstances within the endpoint-associated participant display that are indicative of a face, such as motion or recognition as a face and not as one or more typographic characters) between each endpoint and a facial versus other image that exists in the current meeting’s software. This can inform a display of the extent to which meeting endpoints display faces.

[0419] Among those endpoints that do show faces, artificial intelligence machine recognition principles, once trained, can provide algorithms to assist other real-time recognition tasks. An analysis of facial eye gaze versus likely camera locations near a display , as well as of repeated changes in eye gaze direction, can help determine the extent to which the immediate content may be considered engaging for meeting participants, or to which they exhibit indifference, or even apathy.

[0420] Apathy or indifference are also possible outcomes of a similar Ai-assisted real-time algorithmic analysis of positive (approving, supportive, beneficial) versus negative (disapproving, unpleasant, unwelcome, objectionable) facial attributes.

[0421] Such real-time scoring information may help a presenter adjust pacing, emphasis or other performance attributes to improve the participant acceptance of presented content. Such information, especially when logged with immediate content also indicated in such logs, may also help identify any of several individual or organizational needs (e.g., modified / improved content, training, invited-participant profiles and so on).

[0422] A wide variety of examples / embodiments are presented herein. These may be used in any combination.

Claims

Claims:

1. A see-through display system including: a projector; and a transparent display surface, the display surface including horizontal and / or vertical microlouvers configured to control how much light passes through the display surface and / or how much light is reflected from the display surface.

2. The system of claim 1, wherein the microlouvers are configured to function with a viewer at a specific distance perpendicular from the display service.

3. The system of claim 1 or 2. further comprising an ultra-short throw projector disposed to project a transparent image on the display surface, the ultra-short throw projector optionally being configured to be disposed less than 15, 10, 5, or 2 inches from the display surface.

4. Tire system of any one of the preceding claims, wherein the microlouvers are approximately 15 microns wide, 100-180 microns high, and / or separated by 180-190 microns.

5. The system of any one of tire preceding claims, wherein the ultra-short through projector is characterized by a throw ratio between 0.2 and 0.5.

6. The system of any one of the preceding claims, wherein the ultra-short through projector is configured to operate at an offset of at least 10, 15, or 20%, or any range therebetween, relative to the display surface.

7. Tire system of any one of the preceding claims, wherein the microlouvers are generated using any of the methods described herein.

8. The system of any one of the preceding claims, wherein the display surface includes glass and the microlouvers are bonded directly to the glass.

9. Tire system of any one of the preceding claims, wherein the microlouvers are produced using a combination of both additive and subtractive processes.

10. The system of any one of the preceding claims, wherein the display surface includes reflective beads having a diameter less than 100 microns, e.g., approximately 41 microns, the reflective beads being configured to enhance reflectivity of the display surface.

11. The system of any one of the preceding claims, further including a gap disposed between the display surface and a bezel surrounding tire display surface.

12. The system of any one of the preceding claims, wherein the projector is disposed on a first side of the display surface and a camera is disposed on the other side of the display surface.

13. The system of any one of the preceding claims, wherein the camera or projector is configured to flip a projected or record image to be right reading.

14. Hie system of any one of the preceding claims, wherein the system is configured for a user within the field of view of a camera to point an image presented on the transparent display without turning their back to the camera.

15. The system of any one of the preceding claims, further comprising illuminators disposed on a bezel or frame around the display surface.

Images