Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3 and 12 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Re claim 3, the claim states, “…wherein the material parameter determination neural network is trained by applying the clothing material parameters that receive and output the clothing 3d data of a specific clothing to a rendering engine…” but it is unclear as to what is received and/or the limitation does not make grammatical sense.
Claim 12 is similarly rejected for at least the reasons above.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-2, 8-9, 10-11, 17-18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Meador et al. (US 20230252747) in view of Choi et al. (WO 2022/050810).
Re claim 1, Meador teaches a device for virtual fitting, comprising:
an artificial intelligence server which receives avatar 3D data of a user requesting
virtual fitting ([0052] According to one embodiment, the avatar related data 22 generated by the avatar generation unit 20 can include a three-dimensional mesh indicative of a human shape along with a set of body measurements. The 3D mesh can have a direct correspondence with the user's measurements. That is, the avatar generation unit 20 can generate the mesh from the user measurements or vice versa. The 3D mesh can be to scale with the actual user measurements. Hence, when the simulation engine 24 receives the 3D mesh data, the body measurements can be obtained by directly measuring the mesh. When given the user related measurements, the avatar generation unit 20 can generate the 3D mesh using any existing human-shape modelling technique, such as a Sparse Trained Articulated human body Regressor (STAR), Skinned Multi-Person Linear Model (SMPL), Shape Completion and Animation for People (SCAPE), and the like. Specifically, the avatar generation unit 20 can generate an avatar 22 that highly simulates the body measurements of the user and can be used to virtually “try on” selected garments. As described, the user data 14 can include any combination of images representing the user, biometric data, and information about the user size in known brands. Because of the diversity in the input user data, the avatar generation unit 20 can generate the avatar using multiple different methods or techniques. According to a first method or technique, when only biometric data is available, the avatar generation unit 20 can employ a statistical anthropomorphic analysis technique to infer a full body measurement set from a smaller set (e.g., 3-6 measurements) of body measurements. This can be achieved by using a multi-variable optimization method taking as input the parameters one or more of the foregoing human-shape model techniques, such as STAR, SMPL, SCAPE, and the like, and modifying the human-shape model parameters until the model produces a shape that abides to the measurements given by the user. Then, the rest of the body measurements can be inferred from the obtained shape, as well as from known correlations between human proportions.
clothing 3D data of clothing selected by the user ([0056] The garment rendering system 10 can also include a simulation engine 24 for receiving user selection information, as well as the virtual garment data 18 and the avatar related data 22. The user selection information can include, for example, a selection or indication by the user of a particular garment to view. The virtual garment data 18 can be generated prior to the selection and stored in the system or can be generated by the virtual garment generation unit 16 in real time upon selection by the user based on the garment related data. The simulation engine 24 can drape the virtual garment on the avatar. Specifically, the simulation engine 24 generates a simulation or a simulated garment data 34 that represents how the physical garment looks and drapes on the user's body or avatar 22. As such, the simulation engine employs the avatar related data 22 and the virtual garment data 18. According to the present invention, the simulation engine 24 can overlay one or more virtual garments 18 on the user avatar)
and clothing material parameters for the selected clothing and generates virtual fitting 3D data through neural network operations ([0050] The garment rendering system 10 can also include a virtual garment generation unit 16 for receiving the garment related data 12, such as the garment image data 12A associated with the garment, and then processes the garment image data 12A to generate a virtual garment and associated virtual garment data 18. The virtual garment generation unit 16 can also process the garment textual information 12B to determine information associated with the garment, such as fabric or material information or characteristics, including stiffness, material type, bending resistance, color, and the like. The processed textual information can form part of the virtual garment data 18. The virtual garment data 18 can be a 2D or a 3D virtual representation of the garment. The virtual garment 18 can include datasets that can be processed for the subsequent simulation and rendering of the garments. The virtual garment data 18 can include any selected type of data, including any combination of two-dimensional polygons representing garment patterns, garment images that provide visual details such as prints, textures, look and feel, refraction to light, and the like, fabric or material property data suitable for simulation such as weight, density, stiffness, and the like of the material in the real garment, and fitting information such as a list of sizes (e.g., small, medium, large, extra-large, and the like) and dimensions for each of the listed sizes, such as circumference at the waist, size of the chest area, and the like. The virtual garment data 18 can be conveyed to one or more of the simulation engine 24, the rendering unit 28 and the fit analysis unit 30. According to one embodiment, the virtual garment data 18 can be conveyed to and processed by the simulation engine 24 and can also be conveyed to and processed by the rendering unit 28. The simulation engine 24 can use the virtual garment data 18 (e.g., garment dataset) to simulate how the garment fits or drapes on a user's body. The rendering unit 28 can use the virtual garment data 18 to render the garment by, for example, adding color and texture to the garment when draped on the avatar to create a photorealistic image. Finally, the fit analysis unit 30 can use the virtual garment data 18 to recommend an optimal or proper size of the garment for the user by generating a size recommendation. ([0063] The details of the virtual garment generation unit 16 of the present invention is schematically shown for example in FIGS. 2-6. With specific reference to FIGS. 2 and 5, the illustrated virtual garment generation unit 16 can include a pattern determination unit 40 that is configured to receive garment image data 12A and garment textual data 12B and to generate garment pattern data 42 based on the garment image and the garment textual data. The garment pattern data 42 can be, for example, 2D garment pattern data. The garment pattern data 42 is then received by a mesher unit 44. The mesher unit 44 processes the garment pattern data 42 and generates based thereon three-dimensional garment meshes 45 that are indicative of the garment and thus form at least part of the virtual garment data 18. The mesher unit 44 can employ one or more meshing techniques for subdividing the garment pattern data 42 into a series of geometric spaces or patterns, called meshes or mesh cells, that accurately depict the pattern of the garment. The meshes can form, according to one embodiment, a simplicial complex. The meshes are used as a discrete local approximations of each pattern in the garment. The garment meshes 45 are conveyed to the simulation engine 24 as part of the virtual garment data 18), ([0064] The details of the pattern determination unit 50 are shown for example in FIGS. 3 and 6. The illustrated pattern determination unit 50 can include a classification unit 50 for classifying the garments into one or more garment related classes or categories based on the garment image data 12A and the garment textual data 12B (e.g., multimodal garment data). Specifically, the classification unit 50 can employ any suitable type of classification technique, such as a deep learning model technique, that is suitable for classifying the images and the textual data into classes indicative of different types of garments, such as t-shirts, dresses, trousers, shirts, blazers, coats, and the like. The deep learning model technique can be employed to associate a label or category with every pixel in an image, and can be further employed to recognize a collection of pixels to form distinct categories or classifications. The classification technique can also process different types of data (e.g., multimodal) including garment images and garment textual data. The garment textual data can be converted into one or more word vectors that are embedded in the garment image. According to one embodiment, the word vectors can be embedded in the garment image as a color array. The classification unit 50 then generates garment classification data 52 indicative of the word vectors in the garment image. An example of a suitable deep learning model that can be employed by the classification unit 50 can include a convolutional neural network (CNN) type model. Once the classification unit 50 classifies the garment, the virtual garment generation unit 16 determines the type of garment present in the garment data 12A and 12B)
a rendering server which renders the virtual fitting 3D data using the virtual fitting 3D
data and the clothing material parameters to generate a virtual fitting rendering image ([0057] The system 10 also includes a rendering unit 28 for rendering the avatar data 22 and the virtual garment data 18 via any suitable software application for display on a suitable display device. The rendering unit 28 receives the virtual garment data 18 and the simulated garment data 34 and processes the data to add one or more display or rendering characteristics for enhancing visual realism, such as color, light, texture, and the like, to produce a photorealistic rendered virtual image 32 of the user wearing the simulated garment 34. The rendered virtual image and associated data 32 can be displayed to the end user on a display device. The display device can form part of the system 10, such as for example by forming part of the rendering unit 28, or can be a display device disposed at the user end and hence does not form part of the system 10. The simulation engine 24 and the rendering unit 28 can function as a virtual try-on tool or component of the system. The virtual try-on tool virtually simulates the “wearing” of the garment by the user by overlaying or draping the virtual garment on the user's avatar. This enables the user to determine the overall fit of the garment on the user by assessing the fit of the virtual garment on the avatar. The virtual try-on tool hence delivers custom fit and sizing visualizations to the user)
an image transmission server which transmits pixel data of the virtual fitting rendering
image to the user ([0057] The system 10 also includes a rendering unit 28 for rendering the avatar data 22 and the virtual garment data 18 via any suitable software application for display on a suitable display device. The rendering unit 28 receives the virtual garment data 18 and the simulated garment data 34 and processes the data to add one or more display or rendering characteristics for enhancing visual realism, such as color, light, texture, and the like, to produce a photorealistic rendered virtual image 32 of the user wearing the simulated garment 34. The rendered virtual image and associated data 32 can be displayed to the end user on a display device. The display device can form part of the system 10, such as for example by forming part of the rendering unit 28, or can be a display device disposed at the user end and hence does not form part of the system 10. The simulation engine 24 and the rendering unit 28 can function as a virtual try-on tool or component of the system. The virtual try-on tool virtually simulates the “wearing” of the garment by the user by overlaying or draping the virtual garment on the user's avatar. This enables the user to determine the overall fit of the garment on the user by assessing the fit of the virtual garment on the avatar. The virtual try-on tool hence delivers custom fit and sizing visualizations to the user), and (see [0064] pixels in an image).
and a database which stores user-specific avatar 3D data, clothing-specific 3D data, and
clothing-specific clothing material parameters ([0097] The processor 311 can further be coupled to a database or data storage 380. The data storage 380 can be configured to store information and data relating to various functions and operations of the content characterization systems disclosed herein. For example, as detailed above, the data storage 380 can store information including but not limited to captured information, multimedia, processed information, and characterized content).
wherein the clothing material parameters have a vector form ([0064] The details of the pattern determination unit 50 are shown for example in FIGS. 3 and 6. The illustrated pattern determination unit 50 can include a classification unit 50 for classifying the garments into one or more garment related classes or categories based on the garment image data 12A and the garment textual data 12B (e.g., multimodal garment data). Specifically, the classification unit 50 can employ any suitable type of classification technique, such as a deep learning model technique, that is suitable for classifying the images and the textual data into classes indicative of different types of garments, such as t-shirts, dresses, trousers, shirts, blazers, coats, and the like. The deep learning model technique can be employed to associate a label or category with every pixel in an image, and can be further employed to recognize a collection of pixels to form distinct categories or classifications. The classification technique can also process different types of data (e.g., multimodal) including garment images and garment textual data. The garment textual data can be converted into one or more word vectors that are embedded in the garment image. According to one embodiment, the word vectors can be embedded in the garment image as a color array. The classification unit 50 then generates garment classification data 52 indicative of the word vectors in the garment image. An example of a suitable deep learning model that can be employed by the classification unit 50 can include a convolutional neural network (CNN) type model. Once the classification unit 50 classifies the garment, the virtual garment generation unit 16 determines the type of garment present in the garment data 12A and 12B).
Meador does not explicitly teach wherein the vector form includes a plurality of scalar values.
However, Choi teaches wherein the vector form includes a plurality of scalar values (see p. 5, [41-42], wherein y denotes a vector including six material property parameters and {omega, d} denotes a feature vector) and (see Figs. 3 & 5, wherein scalar values of the clothing material parameters are shown).
Meador and Choi teaches claim 1. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Meador’s virtual fitting system including clothing material parameters to explicitly include clothing material parameters having a vector form including a plurality of scalar values, as taught by Choi, as the references are in the analogous art of modeling clothing materials using parameters having vectors. An advantage of the modification is that it achieves the result having scalar values that represent different attributes of the clothing material parameters, such as seen in Figs. 3 and 5, wherein scalar values represent different material parameters).
Re claim 2, Meador and Choi teaches claim 1. Furthermore, Meador teaches wherein the clothing material parameters are obtained through neural network operations of a material parameter determination neural network that outputs the clothing material parameters using 3D data of the corresponding clothing ([0065] Once the garment classification data 52 is obtained, the pattern determination unit 40 determines the parts present in the garment. To do so, the garment classification data 52 is received and processed by an object detection unit 54 for detecting one or more objects in the classification data. Specifically, the object detection unit 54 can employ one or more object detection techniques for detecting and extracting garment segments or parts from the garment classification data 52. For example, the object detection model detects selected garments parts or segments of each garment that has been classified. The segments that are identified correspond to the type of garment being processed. For example, if the garment is a shirt, then the garment segments include sleeves, necks, sides, and the like. Examples of suitable object detection models can include You Only Look Once (YOLO) such as YOLO 20, a convolutional neural network such as CenterNet ResNet-18 22, a deep neural network such as Inception 18 and Resnet-101 42, and the like. The object detection model employed by the object detection unit 54 can be trained using selected types of training data associated with various types of garments. According to one embodiment, the object detection unit 54 employs a single object detector that can be trained. According to another embodiment, the object detection unit 54 can employ multiple different object detectors. For example, the object detection unit 54 can employ an object detector for each type of garment. The intuition behind training separate object detectors for each garment type is that different garments have different parts or segments. The training of the object detectors can increase the robustness of the overall garment segment detection by removing segment types that are known not to be present in a particular garment (e.g., pant legs in a dress). This increases the ease of re-training as new classes of garments can be trained individually. The object detection unit then generates garment segment data 56), ([0069] The material determination unit 46 can also include a parameter determination unit 74 that receives the garment related data, such as the garment text data, and extracts therefrom or generates based thereon fabric parameters, such as elasticity, density, weight, and the like. According to one embodiment, the parameter determination unit 74 applies to the garment textual data a suitable machine learning technique, such as a natural language processing (NLP) technique, to obtain from the data fabric parameters 48, which are typically provided as part of the garment descriptions, and hence form part of the garment textual data 12B. By simple way of example, a description of a garment can be “made of cotton 80% and spandex 20%.” The properties of the fabric are known and it is now possible to obtain the description using the NLP techniques to map the fabric onto the known simulation parameters), and ([0074] According to another embodiment of the present invention, the tailor unit 80 and the solver unit 84 can be replaced by a machine learning unit, as shown for example in FIG. 7B. The illustrated machine learning unit can employ a one-shot network 92, which employs a deep learning technique, such as a convolutional neural network (CNN). The illustrated one-shot network 92 can be trained in a supervised manner via mapping the pre-simulated garment and the user body information into an already draped garment over the user body. To train the one-shot network 92, a large dataset of garments already draped on avatars present within the simulated garment data 34 are produced using any available method that can produce accurate results. During training, the one-shot network 92 is presented with the already draped garments and the virtual garment data 18 and the avatar data 22 required to produce them. As such, the one-shot network 92 can learn how to map from the data input (e.g., virtual garment data 12 and the avatar data 22) into an output (e.g., the already draped garment on top of the avatar), using a standard supervised learning process. Once the one-shot network 92 is trained, the network is capable of inferring the already draped garment on top of the avatar when given as input the virtual garment data 18 and the avatar data 22. This approach provides the required real-time simulation engine for the virtual try-on aspect of the present invention. Additionally, the entire simulation can be performed in a single pass and does not require multiple solver-tailor iterations, further reducing the time required to simulate the garment).
Re claim 8, Meador and Choi teaches claim 1. Furthermore, Meador teaches wherein the virtual fitting rendering image is displayed on the terminal of the user, and a clothing selection interface and a body selection interface are provided for selecting a clothing or body transformation for transformation of the virtual fitting rendering image (see [0047-0048], user selection include guides for the user to input user data), ([0056] The garment rendering system 10 can also include a simulation engine 24 for receiving user selection information, as well as the virtual garment data 18 and the avatar related data 22. The user selection information can include, for example, a selection or indication by the user of a particular garment to view. The virtual garment data 18 can be generated prior to the selection and stored in the system or can be generated by the virtual garment generation unit 16 in real time upon selection by the user based on the garment related data. The simulation engine 24 can drape the virtual garment on the avatar. Specifically, the simulation engine 24 generates a simulation or a simulated garment data 34 that represents how the physical garment looks and drapes on the user's body or avatar 22. As such, the simulation engine employs the avatar related data 22 and the virtual garment data 18. According to the present invention, the simulation engine 24 can overlay one or more virtual garments 18 on the user avatar), (see [0057], visualization to the user displayed to the end user on a display device).
Re claim 9, Meador and Choi teaches claim 8. Furthermore, Meador teaches wherein request information for image transformation is received from the user terminal, the request information for image transformation includes “coordinate information” for transformation of clothing or body selection information, and the artificial intelligence server and the rendering server generate a virtual fitting transformation image using the selection information and the “coordinate information” for transformation ([0009] The present invention is directed to a garment rendering system for generating a simulated or virtual garment and an avatar of the user, and then overlaying or simulating use of the simulated garment on the user avatar. This allows the user to view the garment on the avatar and hence assess the likely fit of the garment were the user to purchase the garment. More specifically, the garment rendering system of the present invention provides is configured for creating or generating virtual garments from input garment related image data and textual data. This enables the system to generate the virtual garments in multiple ways. Further, the system allows the user to create user avatars using image data uploaded by the user without requiring the user to download a software application) (see [0047-0049], user selection include guides for the user to input user data for acquiring position data of the user (i.e. coordinates information)), ([0056] The garment rendering system 10 can also include a simulation engine 24 for receiving user selection information, as well as the virtual garment data 18 and the avatar related data 22. The user selection information can include, for example, a selection or indication by the user of a particular garment to view. The virtual garment data 18 can be generated prior to the selection and stored in the system or can be generated by the virtual garment generation unit 16 in real time upon selection by the user based on the garment related data. The simulation engine 24 can drape the virtual garment on the avatar. Specifically, the simulation engine 24 generates a simulation or a simulated garment data 34 that represents how the physical garment looks and drapes on the user's body or avatar 22. As such, the simulation engine employs the avatar related data 22 and the virtual garment data 18. According to the present invention, the simulation engine 24 can overlay one or more virtual garments 18 on the user avatar), (see [0057], visualization to the user displayed to the end user on a display device).
Furthermore, to clarify the teachings of coordinate information for transforming of clothing or body selection information and fitting using the coordinate information, Choi explicitly teaches coordinate information for transforming clothing or body selection information (see p. 11, [67-68], x, y, z coordinates for 3d position data).
Claims 10-11 claim limitations in scope to claims 1-2 and is rejected for at least the reasons above.
Claims 17-18 claim limitations in scope to claims 8-9 and is rejected for at least the reasons above.
Claim(s) 3 and 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Meador et al. (US 20230252747) in view of Choi et al. (WO 2022/050810) and Zhao et al. (“M3D-VTON:AMonocular-to-3D Virtual Try-On Network”).
Re claim 3, Meador and Choi teaches claim 2. Furthermore, Meador teaches wherein the material parameter determination neural network is trained by applying the clothing material parameters that receive and output the clothing 3D data of a specific clothing to a rendering engine ([0046] The garment rendering system of the present invention is shown for example in FIG. 1. The garment rendering system can functions as a Virtual Try-On and Size Recommendation (VTON+SR) system. The illustrated garment rendering system 10 is configured to receive any selected type of data from one or more garment related data sources 12, including garment related data, such as garment image data 12A and garment text or textual data 12B. The input garment related data 12 can include any selected type of garment data, from any selected type of input garment data source, including but not limited to any combination of garment related technical packs, garment Computer-Aided Design (CAD) files such as DXF or CLO, three-dimensional scans of a garment, garment image data, textual description of a garment, and the like. The garment related data can include images of garments and associated textual information, including for example one or more of size, color, material or fabric composition, texture, weight, expected fit, garment details and features, and the like. The images of the garments can be provided in any selected file format, such as for example as CAD files. The CAD files can include two-dimensional or three-dimensional images or representations of the garments. The images of the garments can also or alternatively be provided in standard image formats, such as a joint photographic expert group (JPEG) format, a portable network graphic (PNG) format, and the like).
Meador and Choi do not explicitly teach reflecting a loss between the rendered image and a real image of the specific clothing.
However, Zhao teaches reflecting a loss between the rendered image and a real image of the specific clothing (see p. 13242, Fig. 3 showing real clothes images, and wherein during training, the difference between the warped clothing Cw and the ground-truth is used to define warping loss, as seen in equation 3).
Meador, Choi, and Zhao teaches claim 3. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Meador and Choi’s virtual fitting system to explicitly include reflecting a loss between the rendered image and a real image of the specific clothing, as taught by Zhao, as the references are in the analogous art of virtual try-on systems. An advantage of the modification is that it achieves the result of comparing a loss with ground-truth (real image of a specific clothing).
Claim 12 claims limitations in scope to claim 3 and is rejected for at least the reasons above.
Claim(s) 5-6, 14-15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Meador et al. (US 20230252747) in view of Choi et al. (WO 2022/050810) and Liang et al. (US 11869163).
Re claim 5, Meador and Choi teach claim 1. Meador and Choi do not explicitly teach
wherein the artificial intelligence server is trained using physics-based losses.
However, Liang teaches wherein the artificial intelligence server is trained using physics-based losses (see col 3, lines 1-14, physics inspired loss functions used for training).
Meador, Choi, and Liang teaches claim 5. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Meador and Choi’s virtual fitting system to explicitly include physics-based losses training, as taught by Liang, as the references are in the analogous art of virtual try-on systems including material modeling. An advantage of the modification is that it achieves the result of using physics-based losses for warping of materials for virtual fitting of 3d body mesh avatars.
Re claim 6, Meador, Choi, and Liang teaches claim 5. Furthermore, Liang teaches wherein the physics-based losses include strain energy loss that utilizes the physical property of clothing to maintain its original shape, and bending energy loss that utilizes the physical property of minimizing bending of the face of the clothing 3D data (see p. 44-65, stretching, bending energy).
Claims 14-15 claims limitations in scope to claim 5-6 and is rejected for at least the reasons above.
Claim(s) 7, 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Meador et al. (US 20230252747) in view of Choi et al. (WO 2022/050810) and Liang et al. (US 11869163) and DeTemmerman et al. (US 20160163103, hereinafter “De”).
Re claim 7, Meador, Choi, and Liang teach claim 6. Furthermore, Liang teaches wherein the physics=based losses further include collision loss (see col 7, lines 44-65, system optimizes training models and includes penetration loss function to make it collision-aware) and (col 14, lines 50-62, collision avoidance).
Meador, Choi, and Liang do not explicitly teach wherein the loss is to ensure that clothing 3d data does not exist inside avatar data.
However, De teaches wherein the loss is to ensure that clothing 3d data does not exist inside avatar data ([0010] There is then a need for a tool allowing adapting the digital model of a garment to the morphology and size of an avatar. For example, the model of a dress has to be adapted to wearers of different sizes and morphologies. Ideally, if the morphology or size of the avatar is changed, the garment should adapt itself quickly and with minimal or no need for user intervention. In particular, such a software tool should be able to identify and solve “collisions” or “clash”, i.e. situations where some parts of a garment are situated inside the body of the avatar) and (see claim 1, “…detecting collisions between the skin of the model of the avatar and the meshes of the un-deformed model of the garment and, whenever a collision is detected, displacing a vertex of the mesh of the model of the garment away from the skeleton of the model of the avatar along said displacement direction).
Meador, Choi, Liang, and De teaches claim 7. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Meador and Choi and Liang’s virtual fitting system include physics-based losses training including collision loss to avoid collisions, to explicitly includes collisions directed towards data existing inside an avatar, as taught by De, the references are in the analogous art of 3d modeling of garment data on a 3d avatar. An advantage of the modification is that it achieves the result of using collision detection to displace vertex data that collides with 3d mesh data of the avatar, improving the rendering of the virtual fitting system.
Claim 16 claims limitations in scope to claim 7 and is rejected for at least the reasons above.
Allowable Subject Matter
Claims 4 and 13 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Peter Hoang whose telephone number is (571)270-1346. The examiner can normally be reached Monday-Friday 8:00 am - 5:00 pm PST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Hajnik F. Daniel can be reached at (571) 272-7642. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/PETER HOANG/ Primary Examiner, Art Unit 2616
Prosecution Insights