An improved viewing system for media devices

Abstract

The present invention relates to an improved viewing system for increasing viewer safety during the usage of media devices such as televisions by utilizing advanced facial recognition technology, tailored specifically to an individual's needs and safety concerns The system comprises a display unit (100), a power supply unit (160), a plurality of mini-cameras (110a, 110b, 100c) integrated within a bezel frame (105), a facial recognition processing module (120), a user interface module (130), and an alert notification module (140). The system issues visual and audible alerts and autonomously pauses or shuts down the media device if an unsafe distance is detected, enforcing compliance as per the standard safety viewing guidelines set in the system. The system supports personalized profiles for all age groups and unregistered viewers, promoting healthier viewing habits.

Description

[0001] “AN IMPROVED VIEWING SYSTEM FOR MEDIA DEVICES”

[0002] FIELD OF THE INVENTION:

[0003] The present invention relates to an improved viewing system for increasing viewer safety during the usage of media devices such as televisions through facial recognition tool.

[0004] BACKGROUND OF THE INVENTION:

[0005] In numerous households, it is evident that young children and certain vulnerable adults, especially older people, often engage in risky behaviour of watching television from proximity.

[0006] Studies from recent cases indicate that the standard safe viewing distance for any digital media device, including television, varies depending on screen size, resolution, and age. According to standard guidelines, toddlers aged 1-3 years should stay 4-5 feet away from a 32-inch TV. Preschoolers aged 3-5 years should keep a distance of 5-6 feet. School-age children between 6- 12 years should maintain 8- 10 feet from a 40-inch TV. Teenagers aged 13- 18 years and adults should keep 10- 12 feet away from a 55-inch TV. These guidelines promote healthy viewing habits and help to prevent the negative effects of watching TV from proximity.

[0007] However, the habit of watching television in proximity not only poses challenges for parents enforcing safer viewing distances but also raises health concerns, particularly children's eye health. Moreover, watching TV from close distances impacts children's ear health, especially if the decibel levels are excessively high. The prolonged exposure at such close-range leads to several health issues, including eye strain, long-term visual and auditory impairments. The need for a solution that addresses this issue is evident, as parents require reliable, efficient, and easy-to-implement measures that alleviate their concerns and reduce the stress associated with managing their children's television viewing habits. The current technologies offer limited solutions to address these problems effectively. The traditional approaches, such as verbal reminders or setting physical boundaries around the TV area, often prove ineffective. Younger children fail to fully comprehend the risks or forget to maintain a safe distance as instructed by their parents or any guardian. Furthermore, existing child safety products, such as barriers or electronic warning systems, are cumbersome, intrusive, expensive, and not user-friendly to small toddlers and young children. These existing solutions often fail to integrate seamlessly into the living environment and require constant monitoring and manual intervention by the parents, which is both inconvenient and ineffective.

[0008] Thus, there is a need for an improved solution that provides a non-intrusive yet effective means of communication that ensures that the children maintain a safe distance from the television or any digital visual media devices, thereby providing a peaceful environment for parents and safeguarding children's eye health.

[0009] PRIOR ART AND ITS DISADVANTAGES:

[0010] A Chinese patent application CN105872757A, provides a method and an apparatus for maintaining a safe television watching distance, belonging to the technical field of smart televisions. The method includes the steps stating a display screen of a television is started through a body sensing apparatus that monitors the distance between an audience in front of the television and the television in real time; and if the distance is less than a preset safe distance, warning is notified by the invention to remind the audience of a reasonable distance from the television. Through the adoption of the method and the apparatus, the audience in front of the television can be reminded of a safe distance when watching television, and extra hardware devices are not increased therefore, a user is prevented from watching television too close and avoid the eyesight problems.

[0011] Herein, the prior art mainly relies on the body sensing device to detect the proximity of the viewer, but fails to identify specific individuals and / or apply user-specific safety thresholds. Furthermore, the prior art fails to provide an efficient viewing system that allows for the registration and storage of facial profiles for individuals of any age group. Moreover, said prior art does not accommodate the diverse needs and preferences of users, including children and geriatric populations. Additionally, the systems of art lack any advanced recognition technologies, restricting their ability to deliver precise and responsive user experiences. In addition, the prior art is susceptible to false alarms triggered by non-human beings, such as pets or movements resembling those of a human body, leading to inaccurate alerts and unintended interruptions during regular viewing sessions. As a result, the prior art presents a generic and inadequate solution for safeguarding the visual health of all users across various household environments.

[0012] A Chinese patent application CN103295370A discloses a method and system for preventing myopia by monitoring the distance between the eyes and a screen. The method and system for preventing myopia by monitoring the distance between the eyes and the screen is used for detecting the distance between the face of a person and the screen, and said system is also capable of automatically detecting and reminding the user when the distance is too short, and therefore the eyesight of children is effectively protected. According to the technical scheme of cited prior art, the method comprises the steps of collecting images in front of the screen; processing the collected images, identifying key characteristics on the face and calculating parameter values of the key characteristics on the face; comparing the calculated parameter values of the key characteristics on the face with corresponding preset parameter values of the key characteristics, confirming that the distance between the face and the screen is smaller than a preset distance if the parameter values of the key characteristics are larger than the preset values based on the perspective law that the object is bigger when it is near and smaller when it is far, and reminding people of the short distance.

[0013] Herein, the prior art provides a method to estimate facial proximity and issue generic reminders, but fails to incorporate individual facial recognition that stores individual facial profiles and / or supports user- specific viewing thresholds, limiting its ability to deliver personalized safety interventions. Additionally, the prior art mainly focuses on myopia prevention, which may not directly relate to viewing distance across different individuals, since not all close viewing behavior indicates myopia that varies with age and clinical factors. Furthermore, the prior art fails to provide an imaging system integrated into the television frame that ensures complete coverage of the viewing area and fails to operate effectively under various lighting conditions. Furthermore, the prior art fails to incorporate controlled enforcement mechanisms, leading to inconsistent adherence to viewing distance guidelines, particularly among children and other vulnerable users.

[0014] Another Chinese patent application CN 106931923A discloses a kind of sight protection chip. This chip utilizes user image data captured by a camera and the physical distance between the user and the camera to calculate the user's eye distance, eye position, and duration of use, and then provides an analysis of the user's eye health status. In addition to that, said chip is incorporated into various devices such as mobile phones, tablets, PCs, TVs, and gaming consoles either as standalone chips or as part of the primary chip in these devices' subsystems.

[0015] Herein, the prior art mainly focuses on analyzing long-term eye health metrics using image data and usage duration, but fails to consider real-time behavioural enforcement, such as automated alerts or display control, based on unsafe proximity. In addition to that, the sight protection chip disclosed in the prior art for the television needs to be installed in the television hardware, which is costly and requires high maintenance to operate. Also, the prior art fails to provide an imaging system around the television that ensures complete coverage of the viewing area and fails to operate effectively under various lighting conditions. Moreover, the prior art lacks a facial recognition system that stores facial profiles along with customizable safe distance settings for individuals with a real-time monitoring user interface.

[0016] DISADVANTAGES OF THE PRIOR ART:

[0017] The existing prior art suffers from all or any of the following disadvantages: • Most of the prior art fails to provide a facial recognition package that allows for the registration and storage of facial profiles for individuals of any age group, making the prior art less applicable for all age groups.

[0018] • Most of the prior art fails to include user identification, prioritizing only a minimum distance between media devices and viewers, irrespective of their age or physical traits. This leads to a lack of personalized user experience and an over-reliance on a generic distance metric.

[0019] • Most of the prior art fails to provide an imaging system integrated into digital visual media devices, such as a television frame, which prevents complete coverage of the viewing area and effective operation under various lighting conditions.

[0020] • Most of the prior art relies solely on the viewer's experience to act on the visual or auditory safe distance warning instructed by any media device without any further intervention from the system directly.

[0021] • Most of the prior art fails to incorporate any controlled enforcement mechanisms, leading to inconsistent adherence to viewing distance guidelines, particularly among children and other vulnerable users.

[0022] • Most of the prior art fails to integrate with the media device's operational controls and / or other smart home systems, limiting its detection capabilities to only monitoring the presence of the viewer’s distance without any finer recognition system.

[0023] • Most of the prior art fails to provide any specific systems that are tailored specifically to the personalized settings and safety concerns of any age group.

[0024] • Most of the prior arts are susceptible to false alarms triggered by nonhuman beings, such as pets or movements resembling those of a human body, leading to inaccurate alerts and unintended interruptions during regular viewing sessions. • Most of the prior arts are mainly generic and aim to prevent eye disorders like myopia, but they are not directly related to viewing distance among different individuals, since not all close viewing behaviour indicates myopia that varies with age and clinical factors.

[0025] OBJECTS OF THE PRESENT INVENTION:

[0026] The main objective of the present invention is to provide an improved viewing system for media devices.

[0027] Another main object of the present invention is to provide an improved viewing system for media devices that includes a facial recognition system to store presaved facial profiles along with customizable safe distance settings, with a real-time monitoring user interface.

[0028] Another main object of the present invention is to provide an improved viewing system for media devices that ensures compliance with personalized viewing settings with health-oriented habits for children and other vulnerable adult viewers without relying solely on the viewer's discretion.

[0029] Another main object of the present invention is to provide an improved viewing system for media devices that are tailored specifically to the needs and safety concerns of all age groups.

[0030] Another main object of the present invention is to provide an improved viewing system for media devices that incorporates intervening proactive actions integrated by the system after the issuance of visual and / or auditory warnings due to the inconsistent adherence to safe viewing distance guidelines performed by the viewers.

[0031] Another main object of the present invention is to provide an improved viewing system for media devices that improves the efficiency and adaptability of the system through personalized safety protocols.

[0032] Another main object of the present invention is to provide an improved viewing system for media devices that is capable of integration with various smart home systems and enables a seamless and interconnected home environment. Another main object of the present invention is to provide an improved viewing system for media devices that are employed for complete monitoring coverage around the digital media devices on a real-time basis for viewing proximal safe distance behaviours of large individual groups.

[0033] Another main objective of the present invention is to provide an improved viewing system for media devices that obviates the limitations of the existing technology.

[0034] BRIEF DESCRIPTION OF DRAWINGS

[0035] Various other objects, features, and attendant advantages of the present invention will become fully appreciated as the same becomes better understood when considered in conjunction with the accompanying drawings, in which like reference characters designate the same or similar parts throughout the several views, and wherein:

[0036] Figure 1: Shows an internal block diagram that depicts the structural configuration and the interconnections among all the internal components of the present invention.

[0037] Figure 2: Shows a flow chart of a facial recognition processing module according to the present invention.

[0038] Figure 3: Shows a method flow chart of an improved viewing system for media devices according to the present invention

[0039] Figure 4: Shows the working flowchart of the method of the facial recognition processing module according to the present invention.

[0040] Figure 5: Shows a front view of the media display screen that illustrates the preset-coverage zone to maintain a safe viewing distance for the child viewer, according to the present invention.

[0041] Figure 6: Shows the settings menu for configuring the Sentinel Eye feature that allows the facial recognition profiles to be added and saved, according to the present invention. List of reference numerals:

[0042] (100) Display unit

[0043] (110a, 110b, 110c) Three Mini-cameras

[0044] (120) Facial recognition processing module

[0045] (120a) Image capture module

[0046] (120b) Pre-processing module

[0047] (120c) Facial feature extraction module

[0048] (120d) Processing core module

[0049] (120e) Distance calculation module

[0050] (120f) Decision logic controller

[0051] (120g) Data storage module

[0052] (105) Bezel frame

[0053] (160) Power supply unit

[0054] (130) User interface module

[0055] (140) Alarm notification module

[0056] (150) Audio speaker

[0057] (200) Viewer

[0058] (210) Safe viewing distance zone

[0059] (300) Sentinel eye feature menu

[0060] (310) Add new face option button

[0061] (320) Save option button

[0062] (330) Camera settings option button

[0063] (340) User profiles option button

[0064] (350) Safety Settings option button

[0065] (360) System settings option button

[0066] SUMMARY OF THE INVENTION:

[0067] The present invention relates to an improved viewing system for increasing viewer safety during the use of media devices, such as televisions, by utilizing an advanced facial recognition tool tailored specifically to an individual’s needs and safety concerns. The system comprises a display unit, a power supply unit, a plurality of mini-cameras integrated within a bezel frame, a facial recognition processing module, a user interface module, and an alert notification module. The system issues visual and audible alerts and autonomously pauses or shuts down the media device if an unsafe distance is detected, enforcing compliance as per the standard safety viewing guidelines set in the system. The system supports personalized profiles for all age groups and unregistered viewers, promoting healthier viewing habits.

[0068] DETAILED DESCRIPTION OF THE INVENTION:

[0069] The following description is presented to enable any person skilled in the art to make and use the invention. It is provided in the context of a particular application and its requirements. Various modifications to the disclosed embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and applications without departing from the scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and features disclosed herein.

[0070] It is to be understood that the term “comprising” or “comprises” used in the specification and claims refers to the element of the invention which comprises X, Y, and Z, which means that the invention might have other elements in addition to X, Y, and Z. For example, their invention could include A, B, and / or C as long as it also has X, Y, and Z.

[0071] The present invention relates to an improved viewing system for increasing viewer safety during the usage of media devices such as televisions by utilizing advanced facial recognition technology that is tailored specifically to an individual’s needs and safety concerns.

[0072] In some embodiments of the invention, the term “digital media devices” refers to the electronic version of instruments that record, process, transmit, or display digital information. The digital media devices disclosed herein may include screens and monitors of televisions, desktop computers, laptops, gaming systems, and portable tablet devices. Herein, the invention is an integrated safety system that is installed inside digital media devices, including television sets, during their construction and manufacturing stage, especially for newer-generation smart televisions and modern digital display devices.

[0073] In this preferred embodiment of the invention, the internal key components of the system are all built directly into the internal hardware and software architecture of the digital media devices. This arrangement ensures seamless operation, reliable performance, and a user-friendly experience, as all the internal components draw power from the main supply of the digital media devices and are fully accessible through the native interface of the digital media devices using the same remote control or smart menu. Thus, the integrated solution for new digital media devices is inherently more compact, discreet, and robust for the user.

[0074] In another preferred embodiment of the invention, the invention is adapted as an individual external retrofit module for older or non- smart television sets that lack built-in cameras or sufficient processing capability, ensuring broad applicability and user flexibility. In such cases, the improved viewing system is provided as a standalone retrofit unit that includes the internal hardware and software components, which are mounted near the digital media device and connected via a standard interface port, such as HDMI or USB. This modular version enables households with older TVs to benefit from the same intelligent facial-recognition-based safe viewing technology without the need to purchase a completely new television set, resulting in cost savings and reduced electronic waste. Thus, the external module for older digital media devices ensures that the key safety functionality, including advanced facial recognition, personalized viewing distance control, and automated enforcement, is accessible to a wide range of users.

[0075] Therefore, this flexible approach for new and existing digital media devices is intended to promote safe and healthy screen viewing habits and facilitate practical and achievable goals for families, regardless of the type or age of their media device. Accordingly, in some conventional viewing systems, distance monitoring for display screens has been attempted to utilize basic laser, infrared (IR), or other proximity sensors. These conventional systems typically emit beams or pulses configured to detect the presence of any object within the predefined range. Upon detection of such obstruction, the display screen may be configured to pause or stop the operation until the obstruction is removed.

[0076] While such conventional systems offer simple distance sensing, they exhibit significant limitations when compared to the present invention:

[0077] (a) The conventional laser or IR sensors are inherently unable to identify the specific individual positioned near the display screen, as they merely detect the presence of any object, thereby treating all entities uniformly. In contrast, the present invention leverages the use of advanced facial recognition techniques to distinguish between viewers, including different family members, guests, children, or adults. This capability enables the application of personalized safe distance thresholds for each registered user profile that are securely stored in the memory database of the system. For instance, a child viewer is assigned a stricter distance limit compared to an adult viewer with uncompromised eyesight, thereby enhancing personalized eye health protection.

[0078] (b) The conventional proximity sensors are susceptible to generating false positive results, as they may be erroneously triggered by non-human elements such as pets, hand gestures, or reflections, leading to inaccurate system responses. However, the present invention significantly mitigates such errors because its advanced facial recognition techniques are specifically configured to process data for human faces, perform recognition, and actively monitor only relevant human viewing behaviour, thereby substantially improving system accuracy and context awareness.

[0079] (c) Many prior art proximity-based systems typically provide only simple feedback mechanisms, such as an abrupt pause or an undifferentiated auditory signal, upon sensor activation. However, the present invention incorporates a multi-step feedback mechanism managed by the system, including real-time detection with clear on-screen visual warnings that are dynamically tailored to the identified viewer; an audible alert warning if the viewer disregards the visual warning within the predefined duration; and automatic pausing or temporary shutdown of the display screen if the unsafe viewing condition persists.

[0080] Said approach systematically encourages the adoption of safer viewing habits and mitigates user frustration that is often associated with abrupt and non- contextual system interventions.

[0081] (d) Unlike fixed-distance sensors, which operate based on static thresholds, conventional systems lack adaptability to dynamic viewing conditions. The present invention employs trained computational models to exhibit inherent learning and adaptability. This allows the system to adapt to different faces, subtle changes in appearance, or variations in posture over time, ensuring consistent accuracy and reliable performance under dynamic real-world conditions.

[0082] Referring to all the embodiments of the present invention in accordance with Figures 1 and 3, an improved viewing system for media devices comprises:

[0083] - a display unit (100) is configured to display the visual content to the user,

[0084] - a power supply unit (160) is configured to activate and draw power directly from the internal power supply connection of the digital media device,

[0085] - a bezel frame (105) operatively coupled with front and rear frame surfaces of the display unit (100),

[0086] - plurality of mini-cameras (110a, 110b, 110c) integrated into the bezel frame (105) at equidistant edges and configured to capture the real-time images of the designated viewing area upon receiving the fixed triggering signals at 1-2 frames per second,

[0087] - a facial recognition processing module (120) communicatively coupled to the power supply unit (160), configured to receive, process, analyze, calculate, and evaluate the real-time image data from the mini-cameras (110a, 110b, 110c) using advanced computational machine learning tools,

[0088] - a user interface module (130) operatively connected to the facial recognition processing module (120) and configured to manage registration and removal of individual facial recognition profile data, adjust personalized safe viewing distances, and enable and / or disable automatic alert and shutdown features on the display unit (100), and

[0089] - an alert notification module (140) operably connected to the facial recognition processing module (120) and the user interface module (130) and configured to generate audio and visual alerts through integrated audio speakers (150) and the display unit (100) and initiate a pause or power-off function of the media device if a user breaches a predefined safe distance threshold.

[0090] Such components are individually configured and correlated with each other to provide a system that prevents the viewer (200), especially a child or other vulnerable individuals, from watching the media devices, such as television, at a safe viewing distance zone (210).

[0091] In some embodiments of the invention, the media devices, such as a television, are manufactured using standard electronic assembly techniques that incorporate specialized hardware components for facial recognition as per the personalized settings incorporated by the viewer (200).

[0092] In addition to that, the invention disclosed herein implements the display unit (100) as an electronic hardware component for displaying the warning messages and alert notifications when the recognized viewer (200) is too close to the predetermined safe viewing distance zone (210) of the media device.

[0093] In an embodiment of the present invention, the bezel frame (105) is manufactured from a material selected from, but not limited to, plastics, metals (e.g., aluminium alloy), or composite materials, providing structural rigidity and aesthetic finishing to the display unit (100) of the media devices. In this arrangement, the thickness of the bezel frame (105) may exhibit substantial uniformity across its circumference or may vary at specific sections to accommodate integrated internal components.

[0094] The front surface of the bezel frame (105) is substantially co-planar with the active display viewing area of the display unit (100). The corners of the bezel frame (105) are configured with sharp or minimally rounded angles, maintaining the overall rectangular aesthetic of the display media device.

[0095] Continuing to the context of the invention, the improved viewing system comprises plurality of mini-cameras (110a, 110b, 110c). Specifically, three cameras (110a, 110b, 110c) are integrated into the bezel frame (105) and connected directly to the main processing board of the display media devices through the internal wiring harnesses and / or flex cables. Said mini-cameras (110a, 110b, 110c) are positioned at the top-left, top-center, and top-right edges of the display viewing area, respectively, ensuring complete capture from the upper periphery of the media display device.

[0096] Now, referring to Figure 5 of the invention, the mini-cameras (110a, 110b, 110c) having a high-resolution property, continuously capture real-time images across a wide multi-angular range upon receiving triggering signals for 1-2 frames per second, thereby widely monitoring the entire viewing area in front of the digital media device. Herein, the cameras (110a, 110b, 110c) are configured to continuously monitor the viewer

[0200] 's position, for example a child, and capture real-time images at a fixed triggering signal for 1-2 frames per second, relative to a pre-defined safe viewing distance zone (210), which is represented as an arc in front of the display unit (100).

[0097] Additionally, the cameras (110a, 110b, 110c) are capable of capturing high- resolution images suitable for precise facial recognition under various lighting conditions, while accommodating the lateral movement of the child across the entire viewing area. The cameras (110a, 110b, 110c) operate synchronously to deliver a 360-degree field of view, effectively minimizing the blind spots and optimizing the surveillance of the entire viewing area. Furthermore, the miniature cameras (110a, 110b, 110c) continuously capture real-time images of any individual, regardless of their age or physical traits, within the viewing area for 1-2 frames per second.

[0098] Referring now to Figure 1 of the invention, the facial recognition processing module (120) is configured to receive real-time image data simultaneously from the mini-cameras (110a, 110b, 110c). Upon receipt, the facial recognition processing module (120) pre-processes the images to facilitate the identification and tracking of registered users in real-time. This real-time identification and tracking are performed through the application of advanced machine learning tools.

[0099] Further, the facial recognition processing module (120) is adapted to determine the precise distance of the identified viewer (200) from the display unit (100). Based on this determined distance and in comparison with predefined safe distance thresholds stored within the system, the facial recognition processing module (120) generates real-time decision-making alerts and initiates enforcement actions when a viewer (200) breaches a predefined distance.

[0100] Additionally, the facial recognition processing module (120) processes the captured image data to identify facial landmarks. From these landmarks, distinct facial maps are prepared, which are subsequently aligned with registered user profiles securely stored in the system's storage unit.

[0101] In addition to that, the facial recognition processing module (120) allows the registration and storage of facial profiles for each individual user irrespective of their age or their physical traits in the storage unit (not shown in figure), along with customizable and easily operable safe distance settings for each individual through advanced facial recognition tool, along with a real-time monitoring user interface.

[0102] It is to be understood that the term “facial recognition tool” refers to a biometric tool used to match faces in images, usually from photos and video stills, against an existing database of identities. According to all the embodiments of the invention disclosed herein, the facial recognition tool directly integrates with hardware components of the system, creating a proactive health and safety feature that is unique in the home entertainment market.

[0103] Furthermore, the module (120) also integrates with broader smart home systems and / or other connected electronic devices for enhanced automation, ensuring compliance with personalized viewing settings and health-oriented viewing habits for children and other vulnerable adult viewers, without relying solely on the viewer's discretion.

[0104] In the context of the invention, the module (120) analyses and extracts the data from the image captured from the mini-cameras (110a, 110b, 110c) simultaneously, and prepares a facial profile history that is stored in the database based on physical characteristics registered by the viewer along with their predetermined distance threshold.

[0105] Herein, the module (120) is configured to capture and register data corresponding to the facial characteristics and physical features of the viewer (200). The acquired data comprises detailed information related to the geometry, contours, and unique characteristics of the viewer’s face. Said data is subsequently processed and analyzed by the module (120) through the application of advanced computational techniques based on machine learning models. During this processing, the captured facial characteristics and physical features are converted into precise geometrical and numerical values. This transformation facilitates the analysis of spatial relationships between facial landmarks and enables the system to accurately identify and distinguish between individual viewers.

[0106] Referring now to Figure 3 of the invention, the module

[0120] comprises:

[0107] • an image capture module (120a) configured to collect real-time raw image data simultaneously from the plurality of mini-cameras (110a, 110b, 110c) for enabling complete facial mapping; • a pre-processing module (120b) configured to clean and standardize the received raw image data from the image capture module (120a), for performing noise reduction and normalization under varying light conditions;

[0108] • a facial feature extraction module (120c) configured to detect and extract key facial landmarks from the pre-processed image data obtained from the pre-processing module (120b) and generate unique biometric feature vectors for each detected landmark;

[0109] • a processing core module (120d) configured to match said generated biometric feature vectors from the facial feature extraction module (120c) with stored user profiles using trained computational machine learning techniques and continuously adapt for improved recognition accuracy;

[0110] • a distance calculation module (120e) configured to determine a distance of the user from the display unit (100) based on the biometric feature vectors obtained from the processing core module (120d) for distance determination using depth estimation techniques and computational methods;

[0111] • a decision logic controller (1201) configured to evaluate the determined distance against predefined safe distance thresholds calculated by the distance calculation module (120e) and to generate control signals based on the evaluation; and

[0112] • a data storage module (120g) configured to store registered facial profiles securely and their associated user-specific safe distance settings obtained from the decision logic controller (1201) for protecting user privacy.

[0113] Now, again referring to Figure 3 of the invention, the image capture module (120a) directly interfaces with the mini-cameras (110a, 10b, 110c) positioned on the television bezel frame (105) of the system. The image capture module (120a) is configured to collect raw image data simultaneously from multiple angles of the cameras (110a, 10b, 110c), thereby enabling complete facial mapping. Said pre-processing module (120b) is responsible for cleaning and standardizing the raw image data received from the image capture module (120a). In addition, said module (120b) performs various operations, including noise reduction, normalization under varying light conditions, and contrast enhancement to improve the accuracy of subsequent recognition processes executed in the system.

[0114] The facial feature extraction module (120c) detects and extracts key facial landmarks, including eyes, nose, mouth, and jawline, from the pre-processed image data obtained from the pre-processing module (120b) and generates unique biometric feature vectors for each detected landmark.

[0115] The processing core module (120d) houses the trained computational models that match the extracted features from the facial feature extraction module (120c) with stored user profiles. In addition, the processing core module (120d) continuously adapts to improve recognition accuracy by learning from trained computational machine learning techniques and adjusting to subtle changes in facial appearance over time.

[0116] The distance calculation module (120e) calculates the distance between the detected facial features obtained from the processing core module (120d) and the display unit (100). This calculation is performed by analyzing the relative size and spatial position of facial features, utilizing depth estimation techniques and computational methods for spatial distance determination.

[0117] The decision logic controller ( 1201) evaluates whether the viewer is within a safe viewing distance, based on pre-set parameters, and the distance is calculated by the distance calculation unit (120e). Additionally, the decision logic processor (1201) triggers the alert notification module (140) or the automatic display shutdown sequence if the predefined viewing rules are violated.

[0118] The data storage module (120g) securely stores the registered facial profiles and their associated user- specific safe distance settings. Additionally, the data storage module (120g) provides encrypted storage to protect user privacy and ensure data security. Accordingly, the module (120) relies on the advanced computational machine learning techniques specifically adapted for reliable recognition under typical living room conditions, thereby effectively addressing the challenges such as varying light levels, partial facial occlusion, and user movement.

[0119] The implementation of the machine learning capabilities within the facial recognition processing module (120) is as follows:

[0120] (a) employs a Convolutional Neural Network (CNN) model architecture for its computational machine learning techniques,

[0121] (b) pre-trains the CNN model on a large dataset of diverse faces to learn generalized facial feature patterns,

[0122] (c) fine-tunes the CNN model after training for the unique facial profile of each registered user during the setup phase of the system,

[0123] (d) integrates an embedded feedback loop within the processing core module (120d) for enabling the system to continuously refine the model’s accuracy over time by learning from repeated recognition events, thereby ensuring adaptation to subtle changes in the viewer’s appearance,

[0124] (e) incorporates computational logic for distance estimation through the distance calculation unit (120e) that applies depth perception techniques, including monocular depth estimation for utilizing perspective scaling and accurately calculating the real-time distance of the recognized face from the display unit (100),

[0125] (f) optimizes the processing core module (120d) to operate efficiently on the embedded processor and / or a dedicated on-board Al chip of the digital media device, thereby ensuring low-latency performance essential for real-time monitoring,

[0126] (g) processes all captured facial image data on the display unit (100) without being transmitted to any external servers to ensure user privacy, and

[0127] (h) develops the facial recognition processing module (120) to support multiple user profiles, each configurable with different safe distance thresholds to render the system highly adaptable for diverse family viewing habits.

[0128] Thus, the module (120) integrates hardware with a framework of computational models based on CNN architecture to ensure real-time, reliable recognition, precise distance monitoring, and automated enforcement of safe viewing habits, all within a user-friendly and privacy-focused viewing system.

[0129] In a preferred embodiment, the present invention incorporates a specialized image processing module (300), hereinafter referred to as the "Sentinel Eye", associated with the facial recognition module (120) as illustrated in Figure 6. The primary function of the image processing module (300) is to enhance viewer safety by continuously monitoring the viewing distance of a viewer (200) from the display unit (100), under predefined distance guidelines stored within the system's memory.

[0130] Again, referring to Figure 6 of the invention, the module (300) is further configured to provide various selectable options accessible to the user. These options may include, but are not limited to, o adding an 'Add New' button (310) to trigger a facial scan and create a new user profile, and a 'Save' button (320) to store facial data and related safety preferences. o adjusting camera settings (330) and performance behaviour of the display unit (100) that mainly enable / disable individual cameras, adjust frame capture frequency, set recognition sensitivity, and facilitate detection feedback (e.g., show recognition status on screen); o activating or deactivating a user profile setting (340) that mainly interfaces to add, edit, or delete the individual user profiles, wherein said user profiles store facial data, assigned safe distance threshold, and enforcement preferences (e.g., alert only Vs auto power off); o configuring safety settings (350) for safety intervention behaviour that mainly displays warning alert messages or texts with alert duration for 3 seconds, 5 seconds, and 10 seconds, enforcing action after warning by temporarily pausing, dimming, or switching off the media device, and issues a default threshold for unregistered viewers; o and facilitating system settings (360) with general system-level preferences, including reset settings, export logs or usage summaries, enable privacy modes, and storage settings (e.g., image retention duration) o modifying sound settings associated with the audio speakers (150) embedded into the system.

[0131] This module exemplifies the seamless integration of advanced facial recognition technology into user interactions, enabling real-time monitoring and adaptive system responses based on detected user behaviour.

[0132] Again, referring to the preferred embodiment of the invention, the system effectively manages situations where an unknown or unregistered viewer (200) approaches the digital media device, including the display unit (100). When the module (120) detects a face that does not match any of the stored facial profiles within the memory database of the data storage module (120g), said detected face is categorized as "unregistered." In such instances, the system automatically initiates a default safety protocol to safeguard the viewer's eye health and promote safe viewing practices.

[0133] The method for handling registered / unregistered viewers (200) by the system comprises the following steps: a. continuously capturing the viewing area in real time upon receiving triggering signals at every 1-2 frames per second through the minicameras (110a, 110b, 110c), b. transmitting the captured facial data to the facial recognition processing module (120) to pre-process the image data, detect facial landmarks, and generate unique facial maps, c. automatically applying a preset generic safe viewing distance (210), if the detected face fails to correspond to any of the registered user profiles stored in the data storage module (120g). This preset distance is typically established to be equivalent to the default minimum viewing distance recommended for sensitive viewers (200), such as young children. d. allowing uninterrupted viewing, if the registered / unregistered viewers (200) are positioned within the appropriate safe viewing distance zone (210) from the display unit (100), e. initiating the effective protective measures, if the registered / unregistered viewers

[0200] breach this default safe viewing distance zone (210), wherein the protective measures are as follows: i. issuing and displaying an on-screen visual warning and / or an audible alert message, such as "You are too close to the TV. Please move back," on the display unit (100), ii. temporarily pausing or switching off the display unit of the media device, if the viewer (200) fails to respond to the alert within a predefined response time, typically set at five seconds, and iii. restoring the operation only upon the viewer re-establishes the safe viewing distance (210).

[0134] Said default safeguard mechanism in the system ensures that all viewers, irrespective of their registration status, benefit from the protective functions of the invention. Through this mechanism, the parents or guardians assist in preventing any viewer (200), including younger children or guests, from sitting excessively close to the media display device, thereby upholding the invention's core objective of promoting healthier and safer viewing habits for all viewers (200) within the household.

[0135] Now, again referring to the preferred embodiment of the invention, the user interface module (130) is developed as an integrated part of the system, operatively connected to the facial recognition processing module (120) to enable viewers (200) to easily retrieve or remove the facial recognition data from the customizable profile settings for registering facial profiles and establishing safe viewing distance zones (210) based on viewer’s (200) preferences and requirements. Additionally, the user interface module (130) communicates directly with the alert notification module (140) to push visual and / or auditory alert notifications to the display unit (100). The user interface module (130) operates in direct connection with the facial recognition processing module (120), facilitating a range of functionalities, which include:

[0136] • manages the registration and removal of the viewer's facial profiles corresponding to household members;

[0137] • adjusts and configures the personalized safe viewing distances (210) for each registered user profile;

[0138] • enables and / or disables the automatic alert and shutdown features through the alert notification module (140);

[0139] • displays system alerts, warning messages, and any relevant notifications generated from the facial recognition processing module (120); and

[0140] • customizes automatic enforcement actions, such as pausing or deactivating the digital media device by the system, if the safe viewing distance is breached by the viewer (user).

[0141] Said integration ensures that the facial recognition monitoring, user-specific profile settings, and alert response mechanism operate cohesively within the same viewing system, thereby providing an intuitive and user-friendly experience for households.

[0142] In reference to the preferred embodiment of the invention, the alert notification module (140) is configured to independently generate audio and visual alerts through integrated audio speakers (150) and the display unit (100), without the need for any external hardware. This occurs when the facial recognition processing module (120) detects that the viewer is positioned too close to the display unit (100), thereby breaching the predefined safe distance criteria.

[0143] Furthermore, the alert notification module (140) is equipped with the capability to implement proactive enforcement measures, including pausing or deactivating (turning off) the digital media device, after issuing visual and / or auditory warnings due to the viewer’s inconsistent compliance with the established safe viewing distance guidelines. Additionally, the personalization configuration of the alert notification module (140) allows for customizable thresholds and responsive actions based on individual viewer profiles, reflecting their past behavioural viewing following the prescribed safe viewing distance guidelines.

[0144] Accordingly, upon the detection of a user infringing predetermined safe distance proximity thresholds, the alert notification module (140) engages sequentially as follows:

[0145] - displays a visual warning message on the screen, enunciating as “You are too close. Please move back.”;

[0146] - emits an audible signal in the form of a tone or voice alert notification through the audible speakers (150) operably connected to the display media device, if the viewer (200) fails to disengage from the safe distance proximity thresholds incorporated in the system;

[0147] - initiates a pause or power-off function autonomously by the system, if the hazardous condition persists, to mitigate the unsafe scenario,

[0148] - restores the operation only upon the viewer re-establishing the safe viewing distance (210).

[0149] Now, referring to Figure 5 of the invention, the facial recognition processing module (120) is configured to facilitate the registration of the individual facial profiles without imposing any restrictions based on the age, gender, or other demographic factors of the user. Instead, the system relies exclusively on capturing and securely storing the unique biometric facial features of each individual viewer, enabling accurate distinction and identification of one person from another.

[0150] The method for facial criteria and procedures for facial profile registration are as follows:

[0151] • procuring a clear, front-facing image of the individual’s face, ideally captured under controlled lighting conditions, to ensure optimal recognition accuracy by the facial recognition processing module (120), facilitating successful facial registration; • utilizing the configured system to capture and process image data from various angles and expressions during the registration phase, thereby enhancing recognition accuracy under diverse real-world viewing conditions, including but not limited to slight head rotations or varied facial expressions;

[0152] • associating each registered facial profile stored within the data storage unit (120g) with the customizable profile settings, wherein said customizable profile settings may encompass, but are not limited to, specifying a safe viewing distance for the individual and allowing parents or guardians to assign shorter safe viewing distances for younger children or to define particular viewing conditions for users with recognized vision concerns, thus promoting a heightened level of personalized safety;

[0153] • explicitly omitting the analysis or storage of demographic information, such as age, gender, or other generalized physical characteristics, for facial recognition or identification; and

[0154] • mapping and analyzing distinctive patterns of facial features, including the precise distances between facial landmarks and the contours of the face, through the facial recognition processing module (120), wherein these facial features are intrinsically unique to each individual and are independent of general age or gender classifications.

[0155] Said arrangement ensures that the technology is inclusive and adaptable for all family members, enabling parents or guardians to set up profiles for children, adults, or elderly viewers as required, with customizable profile settings specifically tailored to best support each individual’s eye health and overall viewing safety within the household.

[0156] Referring to Figure 2, the preferred embodiments of the present invention provide a method of preparing an improved viewing system for the display media device, as described herein.

[0157] The method for providing an improved viewing system for a media device, wherein the method comprises the steps of: 1. activating (SI 0.1) system through the digital media device power supply unit (160),

[0158] 2. continuously capturing (S10.2) the real-time images of the designated viewing area upon receiving the fixed triggering signals for 1-2 frames per second from the synchronously arranged mini-cameras (110a, 110b, 110c),

[0159] 3. receiving (S10.3), processing, analyzing, calculating, and evaluating the captured real-time image data collected from the mini-cameras (110a, 110b, 110c] for real-time facial landmarks identification of the viewers and their distance from the media display screen using advanced computational machine learning tools through the facial recognition processing module (120),

[0160] 4. retrieving (S10.4), or removing the facial recognition data from the customizable profile settings for registering facial profiles and establishing safe distances based on user preferences and requirements through the user interface module (130),

[0161] 5. triggering (S10.5), the alert notification module (140) when the viewer is too close to the display unit (100) and breaches the personalized safe distance,

[0162] 6. displaying (S10.6), and emitting non-invasive visual and auditory warning contents to the viewer to retreat their steps within the prescribed time level through integrated audio speakers (150) and the display unit (100), and

[0163] 7. pausing (S10.7), or shutting down the digital media device until the viewer has not retreated within the prescribed time level, depending on the personalized settings predefined by the user.

[0164] According to embodiments shown in Figures 4, said facial recognition processing module

[0120] consists of:

[0165] Step 1.1- collecting real-time raw image data simultaneously from the plurality of mini-cameras (110a, 110b, 110c) for enabling complete facial mapping through the image capture module (120a), Step 1.2- cleaning and standardizing the received raw image data from the image capture unit (120a) to perform noise reduction and normalization under varying light conditions through the pre-processing module (120b),

[0166] Step 1.3- detecting and extracting key facial landmarks from the pre- processed image data obtained from the pre-processing unit (120b) and generating unique biometric feature vectors for each detected landmark through the facial feature extraction module (120c),

[0167] Step 1.4- matching said generated biometric feature vectors from the facial feature extraction module (120c) with stored user profiles using trained computational machine learning techniques and continuously adapting for improved recognition accuracy through the processing core module (120d),

[0168] Step 1.5- determining a precise distance of the user from the display unit (100) based on the biometric feature vectors obtained from the processing core module (120d) for spatial distance determination using depth estimation techniques and computational methods through the distance calculation module (120e),

[0169] Step 1.6- evaluating the determined distance against predefined safe distance thresholds calculated by the distance calculation module (120e) and generating control signals based on the evaluation through the decision logic controller ( 120f), and

[0170] Step 1.7- securely storing registered facial profiles and their associated userspecific safe distance settings obtained from the decision logic processor ( 120f) for protecting user privacy through the data storage module (120g).

[0171] The present invention described herein showcases various configurational arrangements primarily formulated in consideration of user preferences and the specific constructional structure of the media display devices. It is, however, contemplated that these arrangements are not limited to the embodiments described herein. Accordingly, the configurational disposition of the elements may vary depending on the structural shape and dimensions of different media display devices, without departing from the scope and spirit of the present invention. Thus, the arrangement shall remain consistent and applicable to all embodiments, modifications, and variations of the described system or method, thereby falling within the scope of the complete specification.

[0172] The implementation of enforced viewing distance is essential for safeguarding viewers, especially children and certain vulnerable adults, from the possible negative effects of being near the television. Thus, the present invention promotes better viewing habits and addresses parental worries about eye health in an unobtrusive yet effective manner.

[0173] WORKING OF THE INVENTION:

[0174] The present invention provides an improved viewing system for media devices designed to enhance viewer safety by utilizing advanced facial recognition technology tailored to individual needs and safety concerns. The system is integrated directly into digital media devices during manufacturing and / or provided as a standalone external retrofit module for existing media devices.

[0175] As illustrated in the accompanying figures 2 and 4, the working of the improved viewing system is described concerning the integrated internal components and their interconnections as follows:

[0176] ■ When the digital media device is powered on, the power supply unit (160) is activated by drawing power to the entire viewing system and initiates its operational sequence.

[0177] ■ The user interface module (130) plays a crucial role in user interaction and system configuration, as it displays various options on the display unit (100), enabling the user (i.e., viewer) to input specific commands. Through this interface, a viewer (200) can:

[0178] - register new facial profiles by capturing their unique biometric features,

[0179] - remove existing facial profiles from the system's memory,

[0180] - adjust personalized safe viewing distances for each registered user, tailoring the system's response to individual preferences, and - enable or disable automatic alert and shutdown features, providing flexibility in enforcement.

[0181] ■ All registered facial profiles and their corresponding user-specific safe distance settings are securely stored in a memory database within the data storage module (120g) of the facial recognition processing module (120).

[0182] ■ The system operates in a continuous real-time monitoring loop to ensure consistent vigilance. This includes-

[0183] The plurality of high-resolution miniature cameras (110a, 110b, 110c), strategically integrated into the bezel frame (105), continuously scan the designated viewing area and capture realtime image data across a wide multi-angular range for precise facial recognition under varying lighting conditions.

[0184] ■ The captured image data is then processed by the facial recognition processing module (120). The facial recognition processing module (120) includes:

[0185] The image capture module (120a) concurrently collects raw image data from all miniature cameras (110a, 110b, 110c) to facilitate comprehensive facial mapping.

[0186] ■ Upon receipt from the image capture module (120a), the raw image data is transmitted to the pre-processing module (120b) that subsequently cleans and standardizes the raw image data by performing operations such as noise reduction, normalization under varying light conditions, and contrast enhancement to improve the accuracy of subsequent recognition processes.

[0187] ■ The processed image data is then sent to the facial feature extraction module (120c) that detects and extracts key facial landmarks from the pre-processed image data and generates unique biometric feature vectors for each detected landmark. ■ The generated biometric feature vectors from the facial feature extraction module (120c) are transmitted to the processing core module (120d) that houses trained computational models, primarily employing a convolutional neural network (CNN) architecture, which matches the newly generated biometric feature vectors with stored user profiles, constituting the real-time identification and tracking of users.

[0188] ■ The processing core module (120d) continuously adapts for improved recognition accuracy by learning from trained computational machine learning techniques and adjusting to subtle changes in facial appearance over time.

[0189] ■ Based on the biometric feature vectors obtained from the processing core module (120d) after the successful facial recognition, the distance calculation module (120e) determines the precise distance of the identified user from the display unit (100). This calculation is performed by analyzing the relative size and spatial position of facial features, using depth estimation techniques and computational methods to determine spatial distance.

[0190] ■ The determined distance from the distance calculation module (120e) is relayed to the decision logic controller ( 1201) that evaluates the determined distance against predefined safe distance thresholds.

[0191] ■ For registered viewers (200), personalized thresholds are stored in the data storage module (120g), and the decision logic controller (1201) evaluates them accordingly.

[0192] ■ For unregistered viewers (200), if the detected face does not correspond to any stored profile, the system optionally applies a preset generic safe viewing distance zone (210), which is then evaluated by the decision logic controller (1201).

[0193] ■ If any viewers (registered or unregistered) breach their respective safe viewing distance zone (210), the decision logic controller (1201) generates control signals to trigger protective measures. ■ A crucial decision point is reached in this stage: “Is the user too close, thereby breaching the threshold?” o If the answer is NO, indicating that a safe viewing distance is maintained, the system continues monitoring, looping back to the "High-resolution Miniature Cameras" for continuous observation. o If the answer is YES, indicating that an unsafe distance has been detected, the system proceeds to the "Alert Notification and Enforcement" phase.

[0194] ■ The control signals from the decision logic controller (120f) activate the alert notification module (140).

[0195] ■ Upon detection of a user infringing predetermined safe distance proximity thresholds, the alert notification module (140) engages sequentially: o a visual warning message, such as "You are too close. Please move back," is prominently displayed on the display unit (100). o If the viewer (200) fails to respond to the visual warning within a predefined response time of five seconds, an audible signal (tone or voice alert) is emitted through the integrated audio speakers (150). o If the hazardous condition persists after both visual and audible warnings, the system autonomously enforces a pause or power-off function of the digital media device to mitigate the unsafe scenario. The operation of the display unit (100) is restored only upon the viewer (200) re-establishing the safe viewing distance zone (210). This tiered approach encourages better viewing habits and promotes compliance.

[0196] ■ The data storage unit (120g) securely stores registered facial profiles and their associated user-specific safe distance settings, providing encrypted storage to protect user privacy and ensure data security. ■ Herein, the user interface module (130) also permits enabling or disabling the automatic alert and shutdown features, displaying system alerts and warning messages, and customizing automatic enforcement actions.

[0197] ■ Once the viewer (200) moves away from the unsafe distance after an alert or pause, the system is designed to either resume normal operation automatically or require manual intervention, depending on the preconfigured user settings. Subsequently, the continuous monitoring loop either restarts or continues seamlessly, ensuring ongoing adherence to safe viewing practices.

[0198] Through this interconnected operation, the system effectively monitors viewing distance, provides timely alerts, and enforces compliance to promote healthier and safer viewing habits for all users, regardless of their registration status or demographic factors.

[0199] ADVANTAGES OF THE INVENTION:

[0200] The present invention of an improved viewing system for media devices provides several advantages over existing viewing systems. These advantages include:

[0201] • tailoring specifically to the needs and safety concerns of any age group,

[0202] • ensuring health-oriented viewing habits for children and vulnerable adults through customizable facial recognition technology on a real-time basis,

[0203] • improving the efficiency and adaptability of the system through personalized safety protocols,

[0204] • offering a comprehensive coverage of the viewing area on a real-time basis for viewing proximal safe distance behaviours of large individual groups,

[0205] • incorporating intervening proactive actions after the issuance of warnings for inconsistent adherence to safe viewing distance, and

[0206] • effectively integrating the system with other connected smart home devices for a seamless and interconnected home environment.

Claims

1. Claim,1. An improved viewing system for media devices for enhancing viewer safety during the usage of media devices, the system comprises:- a display unit (100) is configured to display the visual content to the user,- a power supply unit (160) is configured to activate and draw power directly from the internal power supply connection of the digital media device,- a bezel frame (105) operatively coupled with front and rear frame surfaces of the display unit (100);- plurality of mini-cameras (110a, 110b, 110c) integrated into the bezel frame (105) at equidistant edges and configured to capture the realtime images of the designated viewing area upon receiving the fixed triggering signals at 1-2 frames per second;- a facial recognition processing module (120) communicatively connected to the power supply unit (160), configured to receive, process, analyze, calculate, and evaluate the real-time image data from the cameras [110a, 110b, 110c] using advanced computational machine learning tools, wherein said facial recognition processing module [120] further comprises an image capture module [120a], a pre-processing module [120b], a facial feature extraction module [120c], a processing core module [120d], a distance calculation module [120e], a decision logic controller [120f], and a storage module [120g];- a user interface module (130) operatively connected to the facial recognition processing module (120) and configured to manage registration and removal of individual facial recognition profile data, adjust personalized safe viewing distances, and enable and / or disable automatic alert and shutdown features on the display unit (100), and- an alert notification module (140) operably connected to the facial recognition processing module (120) and the user interface module (130) and configured to generate audio and visual alerts throughintegrated audio speakers (150) and the display unit (100) and initiate a pause or power-off function of the media device if a user breaches a predefined safe distance threshold.

2. The system as claimed in claim 1, wherein said facial recognition processing module [120] is configured to- collect real-time raw image data simultaneously from the plurality of mini-cameras (110a, 110b, 110c) for enabling complete facial mapping through the image capture module (120a),- clean and standardize the received raw image data from the image capture unit (120a) to perform noise reduction and normalization under varying light conditions through the pre-processing module (120b),- detect and extract key facial landmarks from the pre-processed image data obtained from the pre-processing module (120b) and generate unique biometric feature vectors for each detected landmark through the facial feature extraction module (120c),- match the generated biometric feature vectors from the facial feature extraction module (120c) with stored user profiles using trained computational machine learning techniques and continuously adapt for improved recognition accuracy through the processing core module (120d),- determine a precise distance of the user from the display unit (100) based on the biometric feature vectors obtained from the processing core module (120d) for spatial distance determination using depth estimation techniques and computational methods through the distance calculation module (120e),- evaluate the determined distance against predefined safe distance thresholds calculated by the distance calculation module (120e) and generate control signals based on the evaluation through the decision logic controller ( 120f), and- store securely the registered facial profiles and their associated userspecific safe distance settings obtained from the decision logic controller (120f) for protecting user privacy through the data storage module (120g).

3. The system as claimed in claims 1 and 2, wherein said facial recognition processing module (120) implements machine learning capabilities in the following manner:(a) employs a Convolutional Neural Network (CNN) model architecture for its computational machine learning techniques,(b) pre-trains the CNN model on a large dataset of diverse faces to learn generalized facial feature patterns,(c) fine-tunes the CNN model after training for the unique facial profile of each registered user during the setup phase of the system,(d) integrates an embedded feedback loop within the processing core module (120d) for enabling the system to continuously refine the model’s accuracy over time by learning from repeated recognition events,(e) incorporates computational logic for distance estimation through the distance calculation unit (120e) that applies depth perception techniques, including monocular depth estimation for utilizing perspective scaling and accurately calculating the real-time distance of the recognized face from the display unit (100),(f) optimizes the processing core module (120d) to operate efficiently on the embedded processor and / or a dedicated on-board Al chip of the digital media device,(g) processes all captured facial image data on the display unit (100) without being transmitted to any external servers to ensure user privacy, and(h) develops the facial recognition processing module (120) to support multiple user profiles, each configurable with different safe distancethresholds to render the system highly adaptable for diverse family viewing habits.

4. The system as claimed in claim 1 , wherein said media device is configured as an integrated safety system installed during its construction and manufacturing stage, or as an individual external retrofit module for older media devices.

5. The system as claimed in claim 1, wherein said system is configured to apply a preset generic safe viewing distance for an unregistered viewer (200) when a detected face does not correspond to any stored facial profile.

6. The system as claimed in claim 1, wherein said alert notification module (140) engages sequentially as follows:- displays a visual warning message on the screen, enunciating as “You are too close. Please move back.”;- emits an audible signal in the form of a tone or voice alert notification through the audible speakers (150) operably connected to the display media device, if the viewer (200) fails to disengage from the safe distance proximity thresholds incorporated in the system;- initiates a pause or power-off function autonomously by the system, if the hazardous condition persists, to mitigate the unsafe scenario,- restores the operation only upon the viewer re-establishing the safe viewing distance (210).

Images