Drawing production request method and drawing production method, communication terminal and program implementing them

Abstract

Drawing Production Request Method and Drawing Production Method, Communication Terminal and Implementing Program. The invention relates to a drawing production method, a communication terminal, and an implementing program. One object of the invention is a drawing production request method (RDP) implemented by a communication terminal, said request method (RDP) comprising: generating (ICGN) an input command (ic) from an artificial intelligence-generated graphic (GGN), the input command (ic) being generated based on a current neural signal (sn, snj) of a user (U). Thus, the quality of the produced images will be improved for the average user because the input commands will be determined based on the image conceived (either verbally or graphically) by the user. Figure for the abstract: Figure 1

Description

Title of the invention: Drawing production request method and drawing production method, communication terminal and program implementing them. Technical field

[0001] The invention relates to a drawing production method, a communication terminal and a program implementing it. State of the art

[0002] Drawing, whether artistic or technical, remains today the preserve of certain people more gifted than others. This is despite increasingly sophisticated digital tools that now make it possible to produce high-quality images.

[0003] After several years of development of computer-aided design (CAD) software, which gradually made it possible to draw directly on a computer using a keyboard and mouse, computer-aided design enabled both technical drawing (plans, diagrams, etc.) and artistic drawing (posters, video games, animated films, etc.). Then, the use of graphics tablets and later touchscreen tablets, combined with this CAD software, allowed designers to employ more natural drawing techniques, since they again used a pencil: the stylus.

[0004] Recent years have seen the emergence of generative artificial intelligences (AI) such as MidJoumey™ or Dall-E™ whose input commands are formulated in natural language, such as LLM type language models.

[0005] Whether using traditional digital drawing tools (particularly CAD) such as the Adobe™ suite or generative AI, producing a high-quality image still requires knowing how to use them effectively and having access to the right resources. Therefore, despite this evolution in drawing tools, it is still necessary today to call upon a professional designer, graphic artist, or artist, and / or a specialist in writing "prompts" (i.e., input commands for generative AI) to generate a high-quality image, even using graphic AI. Description of the invention

[0006] One of the aims of the present invention is to provide improvements over the prior art.

[0007] An object of the invention is a drawing production request method implemented by a communication terminal, said request method comprising: generate an input command from a graphical generation by artificial intelligence, the input command being generated based on a current neural signal from a user.

[0008] Thus, the quality of the images produced will be improved for the average user because the input commands will be determined according to the image thought of (either verbally or graphically) by the user.

[0009] Advantageously, the query process comprises: receiving a neural signal captured by a direct neural interface worn by the user.

[0010] Advantageously, the input command is generated based on a succession of neural signals from the user, each neural signal corresponding to at least one element of the image to be generated.

[0011] Thus the quality of the images produced will be improved more precisely, that is to say that it will not be improved for a single element of the image but for all the most important elements of the image thought of by the user.

[0012] Advantageously, the sequence of neural signals of the user corresponds to a set of neural signals from among the following: - a succession of neural signals over a predefined time period; - a succession of neural signals interrupted by an end command;

[0013] - a succession of neuronal signals ending with a neuronal signal associated with a final command.

[0014] Thus, the completion of the input command will either be automatic (because it corresponds to a succession of signals during a predefined time period), or controlled in a neural manner, thus avoiding the user having to change the control mode which could generate errors (in particular an input command generated by forgetting the last neural signal or one taking into account a neural signal distinct from the input command).

[0015] Advantageously, the input control function of a neural signal comprises at least one of the following elements: - a word function of a current neural signal from the user thinking about a word; - a word or several words depending on a current neural signal of the user thinking about an image; - an image function of a current neural signal from the user thinking about a word; - an image function of a current neural signal from the user thinking about an image.

[0016] Advantageously, when the user thinks about an image, the generation of the input command includes: - analyze the image associated with the user's current neural signal, the analysis determining a set of words describing the image; - convert the set of words resulting from the analysis into an input command for the AI-generated graphics.

[0017] Thus, the visual thinking of a single image nevertheless allows an image whose quality has been improved by a complex input command.

[0018] Advantageously, the query process comprises: generate a subcommand of the input command by associating with the input command triggering the generation of an image by an artificial intelligence graphics generator a parameterization of at least one element of the image to be generated according to a neural signal from the user.

[0019] Thus, the image quality is improved by taking into account all of the user's thoughts relating to the image to be generated.

[0020] An object of the invention is also a drawing production method implemented by a communication terminal, said drawing production method comprising: - to require the production of a drawing according to the invention by providing an input command for a graphic generation by artificial intelligence; and - generate an image using artificial intelligence based on the input command provided by the drawing production request.

[0021] Advantageously, according to one implementation of the invention, the various steps of the process according to the invention are implemented by a software or computer program, this software comprising software instructions intended to be executed by a data processor of a device forming part of a communication terminal and designed to control the execution of the various steps of this process.

[0022] The invention therefore also relates to a program comprising program code instructions for the execution of the steps of the drawing production request process or the drawing production process according to the invention when said program is executed by a processor.

[0023] This program may use any programming language and be in the form of source code, object code or intermediate code between source code and object code such as in a partially compiled form or in any other desirable form.

[0024] An object of the invention is also a communication terminal comprising: - a drawing production requester capable of generating an input command from a graphics generator by artificial intelligence, the input command being generated according to a neural signal from a user; - a visual reproduction device capable of reproducing the image provided by the graphic generator using artificial intelligence according to the generated input command.

[0025] Advantageously, the communication terminal includes a device for obtaining neural signals from the user.

[0026] Advantageously, the device for obtaining neural signals is one of the following: - a neural signal receiver connected to a direct neural interface worn by the user; - a direct, user-driven neural interface.

[0027] Advantageously, when the signal acquisition device is a user-worn direct neural interface, the direct neural interface has several electrodes arranged on the direct neural interface so as to be positioned relative to predefined areas of the user's skull when wearing the direct neural interface among the following areas: - Broca's area, - visual area, - occipital lobe.

[0028] Advantageously, the communication terminal comprises at least one of the following devices: - a communication interface with the graphics generator using artificial intelligence; - the graphics generator powered by artificial intelligence. Brief description of the drawings

[0029] The features and advantages of the invention will become more apparent upon reading the description, given by way of example, and the related figures which represent:

[0030] [Fig. 1], a simplified diagram of a drawing production request method according to the invention and of a drawing production method according to a first variant of the invention,

[0031] [Fig.2], a simplified diagram of a drawing production process according to a second variant of the invention,

[0032] [Fig.3], a simplified diagram of a communication terminal according to the invention,

[0033] [Fig.4a], a simplified diagram of a first use case of the invention during a simple request,

[0034] [Fig.4b], a simplified diagram of a second use case of the invention during a simple request,

[0035] [Fig. 5a], a simplified diagram of a third use case of the invention during a complex query,

[0036] [Fig.5b], a simplified diagram of a fourth use case of the invention during a complex query. Description of the implementation methods

[0037] Fig. 1 illustrates a simplified diagram of a drawing production request method according to the invention and a drawing production method according to a first variant of the invention.

[0038] The RDP drawing production request process is implemented by a communication terminal. It comprises: generating ICGN, an input command for a graphic generation by artificial intelligence GGN, the input command ic being generated according to a neural signal sn, sn, current of a user U.

[0039] In particular, the RDP request process includes: receive SN_RCV a neuronal signal captured by a direct neural interface BCI worn by the user U.

[0040] In particular, the input command ic is generated based on a succession of neural signals {snjj of the user : ic({ snj]j), each neural signal sn, corresponding to at least one element of the image to be generated e;.

[0041] In particular, the sequence of neural signals of the user {snjj corresponds to a set of neural signals from among the following: - a succession of neural signals during a predefined time period T; - a succession of neural signals interrupted by an end command, in particular an end command from a human-machine interface such as a mouse, a keyboard, etc.;

[0042] - a succession of neuronal signals ending with a neuronal signal sn; associated with a thnd_rq end command.

[0043] In particular, a neuronal signal sn, sn, current according to which the input command ic is generated is a neuronal signal captured when the user U thinks of at least one of the following objects: - a word twd; - a timg image.

[0044] Thus, according to the user U, he will be able to choose the image generation mode with which he is most comfortable in order to improve the quality of the generated image: textual or visual thinking.

[0045] In particular, the input command ic(sn), ic({ sn,}j) function of a neural signal comprises at least one of the following elements: - a word wd function of a current neuronal signal sn of the user U thinking of a word twd; - a word or several words wd, {wdn}n depending on a current neuronal signal sn of the user U thinking about an image timg; - an image img function of a current neuronal signal sn of the user U thinking of a word twd; - an image img function of a current neuronal signal sn of the user U thinking about an image timg.

[0046] In particular, when the user thinks of a timg image, the generation of the ICGN input command includes: - analyze IMG_NZ the image img, img; associated with the current neuronal signal sn of the user U, the analysis IMG_NZ determining a set of words {wdn]n describing the image; - convert the set of words resulting from the analysis into an input command for the AI-generated graphics.

[0047] In particular, the RDP request process includes: generate a subcommand of the input command ic by associating with the input command triggering the generation GGN, GGNm of an image dg by an artificial intelligence graphics generator GGT, GGTm a parameter pr, pri5 prki of at least one element e, e; of the image to be generated as a function of a neural signal snj5 sn,, snki of the user U.

[0048] In particular, the parameterization subcommand includes at least one parameter pr, pr,, prki from among the following: - position of the element on the image pos(e); - image plane in which the element pln(e) is located; - style sty(e); - colour clr(e); - cg(e) color range.

[0049] In the illustrated examples, a single parameter is associated with a neural signal. However, in general, a single neural signal can be used to determine several distinct parameters of the image to be generated, notably from the list above.

[0050] In particular, a direct neural interface carried by a user U of a communication terminal T implementing the RDP request process provides at least one current neural signal sn, snj of the user U captured by the direct neural interface BCI.

[0051] In particular, the RDP request process includes:

[0052] cerebrally command BCGN the graphic generation by artificial intelligence GGN, GGNm.

[0053] In particular, the brain control of graphic generation by artificial intelligence BCGN includes the reception of neuronal signals SN_RCV.

[0054] In particular, the RDP request process includes:

[0055] analyze SN_NZ the neuronal signal sn, snj provided by the direct neuronal interface (BCI) to determine content wd(sn), img(sn), pr(sn), prk(sk), wd^sn;), imgi(sni), priki(snki)ji <ki<i+i pensé par l’utilisateur U.

[0056] In particular, the analysis of neural signals SN_NZ uses a neural signal database BDSN in which the neural signals sn, sn, are associated, prior to the implementation of the RDP query process, with a content wd, img, pr during a learning of the thought of the user U.

[0057] In particular, the SN_NZ neural signal analysis interprets the neural signals sn, sn, by means of a first artificial intelligence device, notably following prior learning to the RDP query process.

[0058] In particular, the brain control of graphic generation by artificial intelligence BCGN includes the analysis of neuronal signals SN_NZ.

[0059] In particular, the reception of neuronal signals SN_RCV provides the neuronal signals sn, snj received from the user's direct neuronal interface to the analysis of neuronal signals SN_NZ.

[0060] In particular, the RDP request process includes:

[0061] check the type ty? of content thought by the user, in particular the type of content determined by the analysis of SN_NZ neural signals.

[0062] In particular, in the case where the content type ty is an image [fi], the query process includes

[0063] analyze IMG_NZ an image img, img; determined as a function of the current neuronal signal sn, snj of the user U, in particular the image determined by the analysis of the neuronal signals SN_NZ.

[0064] In particular, the analyzed image img, img; by the IMG_NZ analysis is an image associated with the current neuronal signal sn of the user U.

[0065] In particular, the IMG_NZ analysis determines a set of words {wdn]n describing the image img, img; determined as a function of the current neuronal signal sn, snj.

[0066] When the user thinks of several words or images to trigger the generation of a GGN, GGNm image, then the RDP request process includes:

[0067] temporarily store TST data relating to a content function of a neuronal signal sn, in particular in a temporary storage device TM, such as a temporary memory, a temporary content database...

[0068] In particular, in the case where the content type ty is an image [fi], the RDP query process temporarily stores TST a set of words {wdn]n describing the image img, imgi determined according to the current neuronal signal sn, snj.

[0069] In particular, in the case where the content type ty is content related to a picture form [f] (in particular a word wd, wd; related to a picture form [f]=[fw] or an image img, imgi related to a picture form [f]=[fi]), the RDP request process temporarily stores TST this content related to the picture form e, ei: e=wd, e=img, e;=wdi, ei=imgi...

[0070] In particular, a set e, ei of data relating to a single piece of content constitutes a description (in particular a textual description: one or more words and / or a description per image) of the shape of an element of an image to be generated. The data relating to the content includes, in particular, the content itself: for example, a word wd, an image img, or descriptive data of the content: for example, a set of words {wdn]n describing an image img as a function of a neural signal sn: ep [wdi,imgf

[0071] In particular, in the case where the content type ty is a parameter data [pr] of an image to be generated, the RDP query process includes: associating PR_CP the parameter data prj, priji with a data or set of stored data relating to e, ei to one and the same content, in particular in the temporary storage device TM.

[0072] Thus, it is no longer just a piece of data or a set of data that will be stored, but a pair or, more generally, a (J+l)-tuple comprising, in addition to a piece of data or a set of stored data relating to one and the same content, one or more parameters prj, pr^ of the element of the image to be generated: (e, pr), (e, {prk}k=1„K)J=K, {(ei,pr;)J=i,

[0073] When the RDP query process uses several neural signals snj to generate the input command ic, then the RDP query process iterates j=j+l through the previous steps for a new neural signal snj. In particular, the RDP query process comprises: Please check if a new neural signal needs to be taken into account in the ICGN generation of the current IC input command.

[0074] In a first variant of the STP? verification, this STP? verification optionally includes a check of the elapsed time between the first neuronal signal snj=i taken into account in the ICGN generation of the current input command ic and the current instant, if the elapsed time is greater than a predefined time period T. Thus, as long as the elapsed time is not greater than this predefined time period [N], then the RDP request process iterates through the previous steps. Otherwise, that is, from If the elapsed time exceeds this predefined time period [Y], then the STP? check determines the end of consideration of new neural signals in the ICGN generation of the current input command ic. Therefore, as soon as the query process includes at least one iteration j=j+1, then the ICGN generation of the current input command ic is a function of a succession of neural signals {sn,}^.

[0075] In a first variant of the STP? check, this STP? check optionally includes a neural command check. In particular, the STP? check determines whether the current neural signal snj corresponds to a command to end the consideration of new neural signals in the ICGN generation of the current input command ic. If it does not [N], in particular when the current neural signal sn corresponds to one or more data points relating to an image to be generated: shape [f] or parameter [pr], then the RDP query process iterates through the preceding steps. Otherwise, that is, whether the current neural signal snj corresponds to a command to end the consideration of new neural signals in the ICGN generation of the current input command ic [Y], then this end command is implemented.

[0076] In particular, as soon as the STP? check determines the end [Y] of taking into account new neural signals in the ICGN generation of the current input command ic, the RDP request process includes: reading temporarily stored data TRD, in particular in the temporary storage device, such as a set of stored data relating to a single content or a (J+1)-tuple comprising, in addition to a data item or a set of stored data relating to a single content, one or more parameters pq, pr^ of the image element to be generated: (e, pr), (e, {prk}k=i...K)j=K, {(c^pr, )}i=i...i, j=i-• •

[0077] In particular, the RDP request process includes

[0078] - convert LG_CNV, LG_CNVm the data relating to an image to be generated e, {e; }i, (e, pr), (e, {pi'k}k i...K)j=K, {(e^pq)}^ . 1>J=1 ... in an input command ic of the graphic generation by artificial intelligence GGN, GGNm.

[0079] In particular, when a single neuronal signal sn is used to determine the data e relating to a single element of the image to be generated: e=wd, e=img, e={wdn]n, the RDP query process does not involve iteration but performs, as soon as this data e is determined (for example, either simply by analyzing the neuronal signal SN_NZ, or after IMG_NZ analysis of an image img resulting from the interpretation of a neuronal signal sn by the analysis of neuronal signals SN_NZ, etc.), the LG_CNV conversion, LG_CNVm of the data relating to the image to be generated e in the input command ic of the graphic generation by artificial intelligence GGN, GGNm.

[0080] In particular, in the case where the RDP query process is configured to generate input commands from several distinct artificial intelligence graphics generations GGNm, then the RDP query process may include: selecting GGT_SLCT an artificial intelligence graphics generation GGNm from among the several available distinct artificial intelligence graphics generations {GGNm]m.

[0081] Optionally, the GGT_SLCT selection is a function of a neuronal signal sn from the user U following the end [Y] of taking into account new neuronal signals in the generation ICGN of the input command ic in progress and / or data relating to the image to be generated e, {eji, (e, pr), (e, {prk}k=1...K)J=K, {(e^pn)}^!...!,M ..., in particular the type of parameterization of the elements of the image to be generated.

[0082] In particular, the LG_CNVm conversion is a function of the selected artificial intelligence graphics generation GGNm.

[0083] The LG_CNV, LG_CNVm conversion is notably carried out on the basis of a language model such as a large language model, also called LLM or "large language model" in English.

[0084] In particular, the generation of ICGN input commands includes one or more of the following steps: - check the type of content intended by the user; - analyze IMG_NZ an image img, img; determined as a function of the current neuronal signal sn, sn, of the user U; - temporarily store TST data relating to a content function of a neuronal signal sn in particular in a temporary storage device TM; - associate PR_CP the parameter data prj, priji with a data or set of stored data relating to e, ei to one and the same content; - Please check if a new neural signal needs to be taken into account in the ICGN generation of the current input command; - read the temporarily stored TRD data; - convert LG_CNV, LG_CNVm the data relating to an image to be generated e, {e;}i, (e, pr), (e, {prk}k=i...K)j=K, {(Ci,pri)}iii.ii ... into an input command ic of the graphics generation by artificial intelligence GGN, GGNm; - select GGT_SLCT a graphic generation by artificial intelligence GGNm from among the several distinct graphic generations by artificial intelligence available {GGNm]m.

[0085] In particular, the brain control of the graphic generation by artificial intelligence BCGN includes one or more of the following steps: - check the type ty? of content thought of by the user; - analyze IMG_NZ an image img, img; determined as a function of the current neuronal signal sn, snj of the user U; - temporarily store TST data relating to a content function of a neuronal signal sn in particular in a temporary storage device TM; - associate PR_CP the parameter data prj, priji with a data or set of stored data relating to e, ei to one and the same content; - Please check if a new neural signal needs to be taken into account in the ICGN generation of the current input command ic; - read the temporarily stored TRD data; - convert LG_CNV, LG_CNVm the data relating to an image to be generated e, {e; }i, (e, pr), (e, {pi'k}k i...K)j=K, [^.^1^(.-,11-1 ... in an input command ic of the graphic generation by artificial intelligence GGN, GGNm ; - select GGT_SLCT a graphic generation by artificial intelligence GGNm from among the several distinct graphic generations by artificial intelligence available {GGNm]m; - generate an ICGN input command from a graphical generation by artificial intelligence GGN, GGNm.

[0086] A particular embodiment of the RDP drawing production request method according to the invention is a program comprising program code instructions for executing the steps of the RDP drawing production request method according to the invention when said program is executed by a processor.

[0087] In particular, a DAP drawing production method is implemented by a communication terminal. Said DAP drawing production method comprises: - requesting RDP to produce a drawing according to the invention by providing an input command ic of a graphic generation by artificial intelligence GN, GNm; and - generate GGN, GGNm by artificial intelligence a dg image according to the input command ic provided by the RDP drawing production request.

[0088] In particular, the generation of an image by artificial intelligence (GGN, GGNm) comprises: - either a generation by local artificial intelligence LGGN, LGGNm, in particular by a graphic generator by artificial intelligence LGGT, LGGTm implemented in the communication terminal T implementing the DAP drawing production process of the dg image,

[0089] - that is: + an emission D_RQ of a request to a graphics generator by artificial intelligence GGT, GGTm remote from the communication terminal T implementing the DAP drawing production process of the image dg, the emitted request d_rq including in particular the generated input command ic; + a D_NSW reception of a response from the GGT, GGTm remote artificial intelligence graphics generator to the d_rq request issued, the response containing the image generated dg by the GGT, GGTm remote artificial intelligence graphics generator.

[0090] In particular, when the AI-generated graphics generation of an image (GGN, GGNm) can use several distinct graphics generators: local (LGGNm) or remote (GGTm), then the GGT_SLCT selection triggers the provision of the generated input command (ic) to the selected graphics generator (GGTm) / graphics generation (LGTm). For example, in the case of a remote graphics generator (GGTm), the issued request (d_rq) is either addressed to the selected remote graphics generator (GGTm), or the issued request (d_rq) includes an identification (ggtm) of the selected remote graphics generator (GGTm).

[0091] Thus, the DAP drawing production process provides a dg image generated according to at least one neural signal from a user, in particular a DRPR visual reproduction device from the communication terminal T.

[0092] A particular embodiment of the DAP drawing production process according to the invention is a program comprising program code instructions for executing the steps of the DAP drawing production process according to the invention when said program is executed by a processor.

[0093] Figure [Fig. 2] illustrates a simplified diagram of a drawing production method according to a second variant of the invention.

[0094] The DAP drawing production process comprises: - require the production of an RDP drawing according to the invention, in particular as illustrated by [Fig. 1] and - generate GGN, GGNm graphically by artificial intelligence an image dg, dgi according to an input command ic provided by the drawing production request.

[0095] In particular, the drawing process comprises: capture SN_CPT a neuronal signal sn, snj5 sn1, sn / from a user U of the communication terminal implementing the drawing production process, the captured neuronal signal sn, sn,. sn, sn,, sn1, sn / being provided to the RDP drawing production request to generate the input command ic.

[0096] In particular, the drawing production process comprises: visually reproduce GRPR the generated dg, dgb image

[0097] In particular, the drawing production process includes a 1=1+1 iteration of drawing production based on an image generated by drawing production during the previous iteration 1.

[0098] For example, to generate a first 1=1 image, the user U thinks of a word twd or an image timg (j=l), the brain of the user U then produces a neuronal signal snHI, snj=i1=1. This first neuronal signal sn1, sn / is captured SN_CPT and provided to the RDP drawing production request.

[0099] Optionally, still to generate this first 1=1 image, user U thinks of one or more other words twd or an image timg (j=2.. .J), the brain of user U then successively produces one or more other neural signals sn / jsn / }. Each jth neural signal sn / is successively captured SN_CPT and provided to the RDP drawing production request.

[0100] The RDP production request generates a first input command ic^i based on a neural signal sn1 or several successive neural signals {sn / }j.

[0101] Next, the graphic generation by artificial intelligence GGN, GGNm generates a first image dgi, according to the first input command ich itself a function of one or more neural signals sn'jsn / jj.

[0102] And, the graphic generation by artificial intelligence GGN, GGNm provides the first generated image dgi to the visual reproduction GRPR. Thus the first generated image dgi is reproduced, for example, on a screen, by an immersive reality device, such as an AR, VR or RX headset or glasses, by a 2D, 3D, holographic projector...

[0103] User U, viewing the first generated image dgi, may want to modify it. Therefore, to generate a new image 1=1+1=2 following this first generated image dgi, user U looks at the first generated image dgi reproduced (j=1). User U's brain then produces a neural signal snj=iH2. This first neural signal sn / is captured by SN_CPT and provided to the RDP drawing production request.

[0104] Then, still to generate this second image 1=2, the user U thinks of one or more other words twd or an image timg (j=2... J) depending on the desired modification(s). The brain of the user U then successively produces one or more other neural signals sn22,{sn / }. Each jth neural signal sn / is successively captured SN_CPT and provided to the RDP drawing production request.

[0105] The RDP production request generates a second input command here=2 based on these successive neural signals {sn / ]j.

[0106] Next, the graphic generation by artificial intelligence GGN, GGNm generates a second image dg! 2 as a function of the second input command ic2, itself a function of these neural signals {sn / jj.

[0107] And, the graphic generation by artificial intelligence GGN, GGNm provides the second generated image dg2 for visual reproduction GRPR. Thus, the second generated image dg2 is reproduced, for example, on a screen, by an immersive reality device, such as an AR, VR or RX headset or glasses, by a 2D, 3D, holographic projector...

[0108] If user U viewing the 1st generated dgi image reproduced wishes to modify it, the above modification steps are repeated as many times as necessary 1=3... L.

[0109] The invention also relates to a medium. The information medium can be any entity or device capable of storing the program. For example, the medium can include a storage means, such as a ROM, for example a CD-ROM or a microelectronic circuit ROM, or a magnetic recording means, for example a floppy disk or a hard disk drive.

[0110] On the other hand, the information medium can be a transmissible medium such as an electrical or optical signal that can be transmitted via an electrical or optical cable, by radio, or by other means. The program according to the invention can, in particular, be downloaded onto a network, especially an Internet-type network.

[0111] Alternatively, the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the process in question.

[0112] In another implementation, the invention is implemented by means of software and / or hardware components. In this context, the term module can refer to either a software component or a hardware component. A software component corresponds to one or more computer programs, one or more subroutines of a program, or more generally to any element of a program or software capable of implementing a function or a set of functions as described above. A hardware component corresponds to any element of a hardware assembly capable of implementing a function or a set of functions.

[0113] In an unillustrated embodiment of the invention, a communication terminal includes a processor capable of implementing the RDP request process and / or the DAP drawing production process according to the invention.

[0114] Figure 3 illustrates a simplified diagram of a communication terminal according to the invention.

[0115] The communication terminal 1 comprises: - a drawing production query 15, 145 capable of generating an input command ic of a graphics generator 140, 4 by artificial intelligence, the input command ic being generated according to a neural signal sn, sn, of a user U; - a visual reproduction device 11 capable of reproducing the image generated dg provided by the graphic generator 140, 4 by artificial intelligence according to the input command generated ic.

[0116] In particular, the communication terminal 1 includes a device 12 for obtaining neural signals from the user U.

[0117] In particular, the neural signal acquisition device 12 is one of the following: - a 12R receptor for neuronal signals sn, snj connected to a direct neuronal interface 2 carried by the user U; - a user-worn 12L direct neural interface U.

[0118] In particular, the direct neural interface 12b 2 comprises one or more electrodes arranged on the direct neural interface 12b 2 so as to be positioned relative to predefined areas of the skull of the user U when wearing the direct neural interface 12b 2 among the following areas: - Broca's area, - visual area, - occipital lobe, - etc.

[0119] In particular, the communication terminal 1 comprises at least one of the following devices: - a communication interface 13 with the graphic generator 4 by artificial intelligence; - the 140 graphic generator by artificial intelligence.

[0120] In particular, a direct neural interface 2, 12, carried by a user U of the communication terminal 1 provides at least one current neural signal sn, snj from the user U captured by the direct neural interface 2, 12b

[0121] In particular, the drawing production requester 15, 145 includes:

[0122] an input command generator 155 for a graphics generator 140, 4 by artificial intelligence. The input command generator 155 is particularly capable of generating an input command for a graphics generator 140, 4 by artificial intelligence.

[0123] In particular, the drawing production requester 15, 145 comprises or consists of:

[0124] A brain controller for a 140, 4 graphics generator using artificial intelligence based on a user's current neural signal. The brain controller for a graphics generator is particularly capable of generating an input command for the 140, 4 graphics generator using artificial intelligence based on a user's current neural signal.

[0125] In particular, the drawing production query 15, 145 includes a neural signal interface 151, for example a neural signal receiver.

[0126] In particular, the drawing production requester 15, 145 includes:

[0127] a device 152 for analyzing neural signals sn, sn, provided by an interface direct neuronal 2,12b the 152 neuronal signal analysis device being capable of determining a content wd(sn), img(sn), pr(sn), prk(sk), wd^snO, imgi(sni), priki(snki)>i <ki <i +i pensé par l’utilisateur u.

[0128] en particulier, le dispositif 152 d’analyse de signaux neuronaux est apte à utiliser une base 153 dans laquelle les sn, sont associés, préalablement la mise en œuvre du requêteur 15, 145 production dessin, un contenu wd, img, pr lors d’un apprentissage pensée

[0129] interpréter au moyen premier d’intelligence artificielle, notamment suite préalable dessin.

[0130] contrôleur cérébral comporte neuronaux.

[0131] l’interface 151 fournit snj="j+i." obtenus, reçus neuronale directe 2, 12, u

[0132] dessin :

[0133] vérificateur 1550 type u, déterminé

[0134]

[0135] analyseur d’image 1551.

[0136] le 1551 analyser des images img; déterminées fonction signal neuronal courant cas où détermine que ty image l’image déterminée

[0137] l’image analysée associée sn

[0138] déterminer ensemble mots {wdn]n décrivant snj.

[0139]

[0140] d’enregistrement temporaire 1554 données relatives stockage 1553, tel qu’une mémoire temporaire, temporaire... stocker temporairement ces lorsque pense plusieurs mots, pour déclencher génération d’une ggn, ggnm.

[0141] snjdans [fi].

[0142] relatif forme e, ei e="img," c^wd,, ce [f]="[fi])," (notamment mot wd; ou relative

[0143] coupleur 1552 associer, donnée paramétrage [pr] générer, prj, priji stocké(e) relative(s) seul et même contenu, 1553.

[0144] ainsi, n’est plus seulement qui sera mais couple manière générale (j+l)-uplet comportant outre {ej; paramètres pq, pr^ l’élément générer (e, pr), {prk}k="i...K)j=K," {(ei,pri)}i="i...i,J=i,"

[0145] lorsque icgn commande d’entrée ic, alors déclencheur d’itération 1555 itérer j="j+l" l’utilisation dispositifs précédents nouveau

[0146] notamment, sn_trg prise compte par exemple, l’obtention snrj+i d’obtention telle capture 121 réception récepteur 12r>j+i.

[0147] In particular, the drawing production query 15, 145 includes: a checker 1555 capable of checking whether a new neural signal should be taken into account in the generation of the current input command ic by the drawing production query 15, 145.

[0148] In particular, the drawing production query 15, 145 comprises: a data reader 1556. The data reader 1556 is capable of reading, as soon as the checker 1555 determines the end of the consideration of new neural signals in the generation of the current input command ic, data temporarily stored, in particular in the temporary storage device 1553, such as a set of stored data relating to a single content or a (J+l)-tuple comprising, in addition to a data item or a set of stored data relating to a single content, one or more parameters pq, pqji of the image element to be generated: (e, pr), (e, {prk}k=i...K)j=K, {(e^pp)J=1...

[0149] In particular, the drawing production requester 15, 145 includes:

[0150] - a converter 1557 capable of converting data relating to an image to be generated e, {eiji, (e, pr), (e, {pi'k}k i...K)j=K, {(e^pq)}^. j ... in an input command ic of a graphics generator by artificial intelligence 140, 4.

[0151] In particular, when a single neural signal sn is used to determine the data e relating to a single element of the image to be generated: e=wd, e=img, e={wdn]n, the drawing production query 15, 145, in particular the iteration trigger 1555, is able to not implement an iteration but provides the data e, as soon as it is determined (for example, either simply by the neural signal analyzer 152, or by the analyzer 1551 of an image img resulting from the interpretation of a neural signal sn by the neural signal analyzer 152, etc.), to the converter 1557 of data relating to the image to be generated e into the input command ic of the graphic generation by artificial intelligence 140, 4.

[0152] In particular, in the case where the drawing production query 15, 145 is configured to generate input commands from several distinct artificial intelligence graphics generators (not shown), then the drawing production query 15, 145 may include: a selector (not shown) of an artificial intelligence graphics generator from among the several distinct artificial intelligence graphics generators available.

[0153] Optionally, the selector is capable of performing the selection based on a neural signal sn from the user U following the completion of the consideration of new signals neural in the generation of the current input command ic and / or data relating to the image to be generated e, {eji, (e, pr), (e, {prk}k=1...K)J=K, {(e^pn)}^!...!,M in particular of the type of parameterization of the elements of the image to be generated.

[0154] In particular, the 1557 converter is capable of implementing a conversion function of the selected artificial intelligence graphics generator.

[0155] In particular, the input control generator 155 of a graphic generator 140, 4 by artificial intelligence includes one or more of the following devices: - the user-thought content type checker 1550 U; - the image analyzer 1551 capable of analyzing an image img, hngi determined according to the current neuronal signal sn, sn, of the user U; - the temporary recording device 1554 capable of temporarily storing data relating to a content function of a neuronal signal sn in particular in a temporary storage device 1553; - the temporary storage device 1553;

[0156] - the coupler 1552 capable of associating the parameter data prj, priji with a data or a set of stored data relating to one and the same content; - the 1555 verifier capable of determining whether a new neural signal should be taken into account in the generation of the current input command ic; - the 1556 data reader capable of reading temporarily stored data; - the 1557 converter capable of converting data relating to an image to be generated e, {eiji, (e, pr), (e, {pi'k}li . k / ik, {fc^prj} ii...ii-, ... in an input command ic of the graphics generator 140, 4 by artificial intelligence; - the selector capable of selecting an AI-powered graphics generator from among several distinct AI-powered graphics generators available.

[0157] In particular, the brain controller (not illustrated) of a graphics generator 140, 4 by artificial intelligence comprises one or more of the following devices: - the neural signal interface 151;

[0158] - the neural signal analyzer 152; - the neural signal base 153;

[0159] - the input control generator 155 of a graphic generator 140, 4 by artificial intelligence;

[0160] - the 1550 user-thought content type checker U; - the image analyzer 1551 capable of analyzing an image img, hngi determined according to the current neuronal signal sn, sn, of the user U; - the temporary recording device 1554 capable of temporarily storing data relating to a content function of a neuronal signal sn in particular in a temporary storage device 1553; - the temporary storage device 1553;

[0161] - the coupler 1552 capable of associating the parameter data prj, priji with a data or a set of stored data relating to one and the same content; - the 1555 verifier capable of determining whether a new neural signal should be taken into account in the generation of the current input command ic; - the 1556 data reader capable of reading temporarily stored data; - the converter 1557 capable of converting the data relating to an image to be generated e, {e;}i, (e, pr), (e, {prk}k=i...K)j=K, {(ei,pri)}i=i...kJ=i ... into an input command ic of the graphic generator 140, 4 by artificial intelligence; - the selector capable of selecting an AI-powered graphics generator from among several distinct AI-powered graphics generators available.

[0162] The drawing production query 15, 145 and / or the input command generator 155 and / or the converter 1557 is capable of providing the input command ic as a function of at least one current neural signal sn, {sn,}j from the user U to the graphics generator 140, 4 by artificial intelligence.

[0163] In particular, the communication interface 13 is capable of relaying the input command ic provided by the drawing production requester 15, 145 and / or the input command generator 155 and / or the converter 1557 to a graphics generator 4 by artificial intelligence separate from the communication terminal 1.

[0164] In particular, the communication interface 13 is capable of relaying the input command ic provided by the drawing production requester 15, 145 and / or the input command generator 155 and / or the converter 1557 to a graphics generator 4 by artificial intelligence remote from the communication terminal 1 via a communication network 3.

[0165] In particular, the communication interface 13 is capable of transmitting the input command ic provided by the drawing production requester 15, 145 and / or the input command generator 155 and / or the converter 1557 to a graphics generator 4 by artificial intelligence remote from the communication terminal 1 via a communication network 3.

[0166] In particular, the communication interface 13 is capable of receiving an image generated dg, dgi by the graphics generator 4 by remote artificial intelligence from the communication terminal 1 via a communication network 3 according to the input command ic previously relayed, issued by the communication interface 13.

[0167] In particular, the communication interface 13 includes a transmitter 130 and a receiver 131. The transmitter 130 is notably capable of transmitting the input command ic provided by the drawing production requester 15, 145 and / or the input command generator 155 and / or the converter 1557 to a graphics generator 4 by artificial intelligence remote from the communication terminal 1 via a communication network 3.

[0168] The receiver 131 is in particular capable of receiving an image generated dg, dgi by the graphic generator 4 by artificial intelligence remote from the communication terminal 1 via a communication network 3 according to the input command ic previously relayed, issued by the transmitter 130.

[0169] In particular, the 140, 4 graphics generator by artificial intelligence is a generative adversarial network or GAN generator.

[0170] In particular, the reproduction device 11 is capable of reproducing the image generated dg, dgi provided by the graphics generator 140, 4 by artificial intelligence, where appropriate via the communication interface 13 and the communication network 3.

[0171] Figures 4a and 4b illustrate two use cases of the invention during a simple query.

[0172] Figure 4a illustrates a simplified diagram of a first use case of the invention during a simple query triggered by the thought of a word.

[0173] User U thinks of a word twd. In our example in [Fig. 4a], this is the word car. The direct neural interface 12b 2 captures the current neural signal sn during (i.e., while) this thought of user U and provides it to the communication terminal 1 according to the invention.

[0174] The communication terminal 1 generates, according to the captured neural signal sn, an input command ic of an artificial intelligence graphics generator which produces following this input command ic a generated image dg.

[0175] The communication terminal 1, in particular the reproduction device 11 of this communication terminal 1, then reproduces the image thus generated dg. In this case, the reproduction device 11 of the communication terminal 1 reproduces an image of a car generated by artificial intelligence.

[0176] Figure 4b illustrates a simplified diagram of a second use case of the invention during a simple query triggered by the thought of an image.

[0177] User U thinks of an image timg. In our example in [Fig. 4b], this is an image of a car, that is, he visualizes any car in his mind. The direct neural interface 12b 2 captures the current neural signal sn during (that is, at (say during) this thought of user U and provides it to the communication terminal 1 according to the invention.

[0178] The communication terminal 1 generates, according to the captured neural signal sn, an input command ic of an artificial intelligence graphics generator which produces following this input command ic a generated image dg.

[0179] The communication terminal 1, in particular the reproduction device 11 of this communication terminal 1, then reproduces the image thus generated dg. In this case, the reproduction device 11 of the communication terminal 1 reproduces an image of a car generated by artificial intelligence.

[0180] Figures 5a and 5b illustrate two use cases of the invention during a complex query.

[0181] Fig. 5a illustrates a simplified diagram of a third use case of the invention during a complex query triggered by several successive thoughts relating to the same element of an image to be generated.

[0182] Initially, user U thinks of a first image timgi. In our example in [Fig. 5a], this is an image of a car, that is, they visualize any car in their mind. The direct neural interface 12b 2 captures the current neural signal during (i.e., while) this thought of user U and provides it to the communication terminal 1 according to the invention as the first neural signal sni related to the image to be generated.

[0183] The communication terminal 1 stores the image and / or words describing the image determined according to the first neuronal signal captured sni as the first image thought timgi by the user U.

[0184] Then, in a second step t2, user U thinks of a second image timg2. In our example in [Fig. 5a], this is a colored image (in this example, the color green), that is, they visualize this color in their mind. The direct neural interface 12b2 captures the current neural signal sn during (that is, while) this thought of user U and provides it to the communication terminal 1 as a second neural signal sn2 relating to the image to be generated.

[0185] The communication terminal 1 stores a color parameter (determined based on the second captured neuronal signal sn2 as the second thought image timg2 by the user U) in association with the previously stored data, in this case the image and / or words describing the image determined based on the first captured neuronal signal sni as the first thought image timgi by the user U.

[0186] And, in a third step t3, user U has a third thought. In this case, he thinks of a word twd3. In our example in [Fig. 5a], it is from a word describing a style: sty (in this example, graffiti style). The direct neural interface 12b 2 captures the current neural signal sn during (i.e., while) this thought of the user U and provides it to the communication terminal 1 as a third neural signal sn3 relating to the image to be generated.

[0187] The communication terminal 1 stores a style parameter (determined based on the third captured neural signal sn3 as the third thought twd3 of user U) in association with the previously stored data, in this case the image and / or words describing the image determined based on the first captured neural signal sni as the first thought image timgi by user U, and the color parameter.

[0188] In an unillustrated example, a single neural signal is used instead of the second and third neural signals to determine both the color and style of the image to be generated.

[0189] Finally, in a fourth step:

[0190] The communication terminal 1 generates, based on the stored data (in this case the image and / or words describing the determined image, i.e., a car in this example, based on the first neural signal captured sni as the first image thought of timgi by the user U; and the associated parameters in our example, namely the color parameter - in this example: green - determined based on the second neural signal captured sn2 as the second image thought of timg 2 by the user U, and the style parameter - in this example: graffiti - determined based on the third neural signal captured sn3 as the third thought of user U), an input command ic from an artificial intelligence graphics generator which, following this input command ic, produces a generated image dg. The input command ic is, in particular, a "prompt" such as: "green car in graffiti style".

[0191] The communication terminal 1, in particular the reproduction device 11 of this communication terminal 1, then reproduces the image thus generated dg. In this case, the reproduction device 11 of the communication terminal 1 reproduces an image of a green car in graffiti style generated by artificial intelligence.

[0192] Fig. 5b illustrates a simplified diagram of a fourth use case of the invention during a complex query triggered by several successive thoughts relating to distinct elements of an image to be generated.

[0193] Initially, user U thinks of a first image timgi. In our example in [Fig. 5b], this is an image of a car, that is, they visualize any car in their mind. The direct neural interface 12b 2 captures the current neuronal signal during (i.e. during) this thought of the user U and provides it to the communication terminal 1 according to the invention as the first neuronal signal sni relative to the image to be generated.

[0194] The communication terminal 1 stores the image and / or words describing the image determined according to the first neuronal signal captured sni as the first image thought timgi by the user U.

[0195] Then, in a second step t2, the user U thinks of a second image timg2. In our example in [Fig. 5b], this is the plane of the image in which the previous shape must be positioned (in this example, the foreground), that is, they visualize this position in their mind. The direct neural interface 12b2 captures the current neural signal sn during (that is, while) this thought of the user U and provides it to the communication terminal 1 as a second neural signal sn2 relating to the image to be generated.

[0196] The communication terminal 1 stores the plan parameter (determined based on the second captured neuronal signal sn2 as the second thought image timg2 by the user U) in association with the previously stored data, in this case the image and / or words describing the image determined based on the first captured neuronal signal sni as the first thought image timgi by the user U.

[0197] Then, in a third step t3, user U has a third thought. In this case, they think of a third image timg3. In our example in [Fig. 5b], this is an image of the sea; that is, they visualize any sea in their mind. The direct neural interface 12b 2 captures the current neural signal sn during (i.e., while) this thought of user U and provides it to the communication terminal 1 as a third neural signal sn3 relating to the image to be generated.

[0198] The communication terminal 1 stores the image and / or words describing the image determined according to the third neuronal signal captured sni as the third thought image timg3 by the user U.

[0199] And, in a fourth step t4, the user U thinks of a fourth image timg4. In our example in [Fig. 5b], this is the plane of the image in which the previous shape, in this case the sea, must be positioned in the image (in this example, background), that is, they visualize this position in their mind. The direct neural interface 12b 2 captures the current neural signal sn during (that is, while) this thought of the user U and provides it to the communication terminal 1 as a fourth neural signal sn4 relating to the image to be generated.

[0200] The communication terminal 1 stores the plan parameter (determined based on the fourth captured neuronal signal sn4 as the fourth thought image timg 4 by user U) in association with previously stored data, in this case the image and / or words describing the image determined according to the third captured neuronal signal sn3 as the third thought image timg3 by user U.

[0201] Finally, in a fifth step t5:

[0202] The communication terminal 1 generates, based on the stored data (in this case, the images and / or words describing the determined images, i.e., a car and the sea in this example, based respectively on the first and third neural signals captured sni and sn3 as the first and third images thought of timgi and timg3 by the user U; and the associated parameters in our example, namely the respective shot parameters - in this example: foreground for the car and background for the sea - determined respectively based on the second and fourth neural signals captured sn2 and sn4 as the second and fourth images thought of timg2 and timg4 by the user U), an input command ic from an artificial intelligence graphics generator which, following this input command ic, produces a generated image dg. The input command ic is in particular a "prompt" such as: "car in the foreground and sea in the background".

[0203] The communication terminal 1, in particular the reproduction device 11 of this communication terminal 1, then reproduces the image thus generated dg. In this case, the reproduction device 11 of the communication terminal 1 reproduces an image of a car in front of the sea generated by artificial intelligence.

[0204] The invention makes it possible, in particular, to generate by thought: - Simple drawings with a single shape: primitive shapes (circle, square, sphere, cube...) or complex shapes (car, house, object, etc.) - complex drawings (landscapes, portraits, posters...) using colors such as primary colors, and / or complementary colors and / or colors from among 10, 50, 256... colors with particular styles (comics, science fiction, painting, photorealistic...).

[0205] Optionally, the invention allows for the generation of an image, in particular a scene, from the user's thoughts (or thoughts) containing a standardized object such as an 8mm bolt, an E27 light bulb, a referenced piece of furniture, etc. The generated image is, in particular, an industrial drawing or an assembly plan, or even a 2D or 3D interior design plan...

[0206] The neural signal analyzer may optionally draw from a database of 2D or 3D drawings to provide the thought image enabling the artificial intelligence generator to control the image of the scene.

[0207] This invention thus makes it possible to draw by thought thanks to a direct neural interface, in particular based on the user's EEG or electroencephalogram.

[0208] This invention is a design tool accessible to a wider audience than just professionals or experienced amateurs for image creation. It also allows the generation of sketches from a simple idea.

[0209] "A picture is worth a thousand words." This saying will be all the easier to apply. to be verified by using the invention during communication.

[0210] Finally, since the control is neural, the invention allows a disabled user access to image generation by artificial intelligence.

Claims

Demands

1. A drawing production request method implemented by a communication terminal, said request method comprising: generating an input command of a graphic generation by artificial intelligence, the input command being generated based on a current neural signal of a user.

2. A query method according to the preceding claim, wherein the query method comprises: receiving a neural signal captured by a direct neural interface worn by the user.

3. A query method according to any one of the preceding claims, wherein the input command is generated based on a succession of neural signals from the user, each neural signal corresponding to at least one element of the image to be generated.

4. A query method according to the preceding claim, wherein the sequence of user neural signals corresponds to a set of neural signals among the following: - a sequence of neural signals during a predefined time period; - a sequence of neural signals interrupted by an end command; - a sequence of neural signals ending with a neural signal associated with an end command.

5. A query method according to any one of the preceding claims, wherein the input command, which is a function of a neural signal, comprises at least one of the following: - a word, which is a function of a current neural signal of the user thinking about a word; - a word or several words, which is a function of a current neural signal of the user thinking about an image; - an image, which is a function of a current neural signal of the user thinking about a word; - an image, which is a function of a current neural signal of the user thinking about an image.

6. A query method according to any one of the preceding claims, wherein, when the user thinks of an image, the generation of the input command comprises: - analyze the image associated with the user's current neural signal, the analysis determining a set of words describing the image; - convert the resulting set of words from the analysis into an input command for the AI-generated graphics.

7. A query method according to any one of the preceding claims, wherein the query method comprises: generating a subcommand of the input command by associating with the input command triggering the generation of an image by an artificial intelligence graphics generator a parameterization of at least one element of the image to be generated according to a neural signal from the user.

8. A drawing production method implemented by a communication terminal, said drawing production method comprising: - requesting a drawing production according to one of the preceding claims by providing an input command for graphic generation by artificial intelligence; and - generating by artificial intelligence an image according to the input command provided by the drawing production request.

9. Program comprising program code instructions for executing the steps of the drawing production request process according to any one of claims 1 to 7 and / or the drawing production process according to the preceding claim when said program is executed by a processor.

10. Communication terminal comprising: - a drawing production requester capable of generating an input command from a graphics generator by artificial intelligence, the input command being generated according to a neural signal from a user; - a visual reproduction device capable of reproducing the generated image provided by the graphics generator by artificial intelligence according to the generated input command.

11. Communication terminal according to the preceding claim, wherein the communication terminal includes a device for obtaining neural signals from the user.

12. Communication terminal according to the preceding claim, wherein the neural signal acquisition device is one of the following: - a neural signal receiver connected to a user-worn direct neural interface; - a user-worn direct neural interface.

13. Communication terminal according to the preceding claim, wherein, where the signal-gathering device is a user-worn direct neural interface, the direct neural interface comprises several electrodes arranged on the direct neural interface so as to be positioned relative to predefined areas of the user's skull when wearing the direct neural interface among the following areas: - Broca's area, - visual area, - occipital lobe.

14. Communication terminal according to any one of claims 10 to 13, wherein the communication terminal comprises at least one of the following: - a communication interface with the AI ​​graphics generator; - the AI ​​graphics generator.

Images