EP4073765A1 - Procédé et système de génération de modèles numériques 3d - Google Patents
Procédé et système de génération de modèles numériques 3dInfo
- Publication number
- EP4073765A1 EP4073765A1 EP20845411.6A EP20845411A EP4073765A1 EP 4073765 A1 EP4073765 A1 EP 4073765A1 EP 20845411 A EP20845411 A EP 20845411A EP 4073765 A1 EP4073765 A1 EP 4073765A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- model
- dimensional
- dimensional digital
- image data
- models
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three-dimensional [3D] modelling for computer graphics
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—Three-dimensional [3D] image rendering
- G06T15/04—Texture mapping
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating three-dimensional [3D] models or images for computer graphics
- G06T19/20—Editing of three-dimensional [3D] images, e.g. changing shapes or colours, aligning objects or positioning parts
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2219/00—Indexing scheme for manipulating 3D models or images for computer graphics
- G06T2219/20—Indexing scheme for editing of 3D models
- G06T2219/2021—Shape modification
Definitions
- the invention relates generally to the generation of three-dimensional digital models, called "3D models”. More particularly, the invention relates to a method and a system for generating three-dimensional digital models.
- 3D modeling is of great interest to understand an object holistically and improve understanding of it.
- a 3D model advantageously replaces a large number of two-dimensional images, called 2D images ”.
- the 3D model can easily be manipulated to view the object from different angles.
- the 3D model also allows for an animated presentation of the object by calculating a plurality of images which are displayed successively, typically at a rate of at least 25 frames per second.
- the 3D model is optimized for real-time calculation of the images, which are rendered directly on the display device.
- This technology provides a high level of user interactivity and animation control by the user, allowing for increased immersion possibilities and a better user experience.
- 3D modeling finds an important application in electronic commerce, known as “e-commerce”, and has experienced strong development there in recent years.
- 3D visualization of products on merchant sites improves understanding of products by customers.
- the customer can interact with the product, for example, by rotating it or zooming in on details, which increases their level of engagement towards an act of purchase.
- Animations using 3D models can easily be integrated into the merchant site, for example, to highlight products or attract the attention of the customer.
- 3D models are created manually with 3D modeling software such as “Blender®”, “3Ds Max®”, “Maya®” and others.
- the manual creation with software of a good quality 3D model is an activity that can require a lot of work.
- a three-dimensional modeling method described by document EP2381421A2 has been proposed.
- This method is computer implemented and is designed to automatically generate a three dimensional model of a product from two dimensional image data.
- shape information is extracted from the two-dimensional image data, as well as a class of three-dimensional shapes to which the represented product belongs.
- a three-dimensional graphic outline of the product is then determined from the extracted information.
- a mesh generator generates a three-dimensional mesh based on the determined three-dimensional graphic outline and the three-dimensional shape class of the product.
- the three-dimensional model of the product is obtained by mapping images to the three-dimensional mesh.
- the images projected on the three-dimensional mesh are extracted from the two-dimensional image data of the product.
- Document US2019147221A1 describes a method for selecting a three-dimensional digital model from an image provided as an input. An object is identified in the input image and one or more candidate three-dimensional models for representation of the object are preselected from a database. A candidate three-dimensional model is finally selected to represent the object in the input image.
- the objective of the present invention is to provide an improved method for generating three-dimensional digital models, allowing easy and low-cost generation of three-dimensional digital models which are suitable, in particular, for real-time applications on Internet sites.
- the invention relates to a method, implemented by computer, for generating three-dimensional digital models in a system for generating three-dimensional digital models for objects represented in two-dimensional images, the method comprising the steps of: a ) assembling a set of diversified data comprising two-dimensional image data and a database of three-dimensional digital models, relating to a plurality of objects; b) training of a multidimensional type convolutional neural network with the set of diversified data so as to obtain a neural model resulting from the training; c) receiving incoming two-dimensional image data representing the object; d) indexing in the database of three-dimensional digital models, using the neural model and a choice function, of at least one three-dimensional digital model having at least one characteristic in common with the object represented by the data incoming two-dimensional image data, the indexing resulting from classification of the incoming two-dimensional image data into different object classes by a function of features recognized in the incoming two-dimensional image data using the neural model, the recognized features including at least one shape feature
- the choice function is based on recognition probabilities of the characteristics in the incoming two-dimensional image data.
- the database of three-dimensional digital models comprises textured and / or non-textured three-dimensional digital models.
- the method also comprises a step of: h) enriching the set of diversified data with the three-dimensional digital model provided in step g).
- the method also comprises a step of: i) three-dimensional visualization of a plurality of three-dimensional digital models indexed in step d), and manual selection of one of the three-dimensional digital models displayed as a model three-dimensional digital image of the object represented by the incoming two-dimensional image data.
- the method also comprises a step of: j) automatic selection, based on the choice function, of a three-dimensional digital model from among a plurality of three-dimensional digital models indexed to step d) as a three-dimensional digital model of the object represented by the incoming two-dimensional image data.
- the invention also relates to a system for generating three-dimensional digital models comprising at least one computer server and a data storage device associated with the computer server, the computer server being connected to a data communication network and allowing user access to the system.
- the system comprises additional means for implementing the method of generating three-dimensional digital models described briefly above, these additional means comprising a convolutional neural network.
- the system for generating three-dimensional digital models of the invention can be produced, for example, in the form of a system accessible in so-called "SAAS” mode (for "Software As a Service” in English).
- SAAS Software As a Service
- the invention also relates to a computer program comprising program code instructions implementing the method for generating three-dimensional digital models described briefly above when these are executed by a device processor. computer science.
- FIG. 1 shows six simplified 2D images of products having different shapes, materials and textures.
- FIG. 2 is a simplified general view of a particular embodiment of the system for generating 3D models according to the invention.
- FIG. 3 is a block diagram showing different steps included in the process for generating 3D models according to the invention.
- FIG. 4 is a block diagram illustrating a multiclassification of incoming 2D images performed in the method for generating 3D models according to the invention.
- FIG. 5 is a block diagram illustrating the emission of suggestions for 3D models and of treatments to be carried out on the 3D models in the system for generating 3D models according to the invention.
- FIG. 6 is a block diagram illustrating the suggestion of a 3D model by the artificial intelligence process of the system for generating 3D models according to the invention.
- FIG. 7 shows a choice function used by the artificial intelligence process of the 3D model generation system according to the invention.
- the objects represented in the images are characterized primarily by shapes, materials, textures and visual effects.
- Material is defined as the type of material that goes into the construction of an object. In the design of a 3D object, the material is largely responsible for the visual rendering of the object. The material has special properties and is therefore a distinctive feature that helps to distinguish objects.
- the materials can be classified arbitrarily, for example, into three main categories, namely, a class called metals, a class called ceramic-type materials and a class called polymers.
- a class called metals will be included, for example, iron, steel, copper and aluminum.
- ceramic-type materials will be included, for example, ceramic and glass.
- polymers will be included, for example, wood, cardboard, plastics, leather and rubber.
- material editors usually allow you to create standard materials with adjustable parameters.
- a program or script for adding material to a 3D model can control the parameters of a standard material to achieve desired visual effects.
- the materials editor contained in a software application of Unity3D® type authorizes the adjustment of parameters on a standard material, such as the parameters designated “generic”, “reflectivity” and “transparency”.
- the "generic” parameter is used to adjust the overall appearance of the object such as its color for example.
- the "reflectivity” parameter allows you to adjust the reflections of the object, for example, to obtain a mat material (diffuse reflection) or a mirror effect (specular reflection).
- the "transparency” parameter allows you to adjust the transparency of the object.
- a texture is a 2D image that is applied to a 3D model, usually after the material is applied, to represent surface details.
- a texture is for example the 2D image of a logo or a brand label on an object.
- the operation of adding texture is called "texture mapping" by those skilled in the art.
- textures can be applied to a material to obtain a realistic 3D model reproducing the object in a fine way. This is then referred to as “texture multimapping” or “multimapping” in English.
- Fig. 1 shows six simplified 2D images, labeled PC1 to PC6, respectively representing a bottle of cider, a can of beer, a bottle of oil, a bottle of soda, a can of paint and a packet of cereal.
- a cylindrical shape characterizes the cider bottle in picture PC1, the beer can in picture PC2, the oil bottle in picture PC3, the soda bottle in picture PC4 and the paint can of the PC5 image.
- the package of cereals in image PC6 is characterized by a parallelepipedal shape.
- a glass material with high specular reflection characterizes the cider bottle in image PC1.
- a metal material, with high specular reflection characterizes the beer can in picture PC2.
- a plastic material, with a medium specular reflection characterizes the oil bottle in picture PC3.
- a plastic material, with a medium specular reflection characterizes the soda bottle in picture PC4.
- a metal material, with very low reflection of matt material characterizes the paint bucket in picture PC5.
- a cardboard material, with very low reflection of matt material characterizes the cereal package in picture PC6.
- An image texture characterizes the cider bottle in picture PC1, the beer can in picture PC2, the oil bottle in picture PC3, soda bottle in picture PC4, the paint pot in picture PC5 and cereal packet in picture PC6.
- the system 1 according to the invention for the generation of 3D models is deployed via an IP data communication network, such as the Internet network, and uses hardware and software resources accessible via this network. network.
- IP data communication network such as the Internet network
- the system 1 uses software and hardware resources available from a cloud computing service provider CSP called “cloud service provider” in English.
- System 1 uses at least one computer server SRC from cloud computing service provider CSP.
- the computer server SRC comprises in particular a processor PROC which communicates with a data storage device HD, typically dedicated to the system 1, and conventional hardware devices such as network interfaces NI and other devices (not shown).
- the processor PROC comprises one or more central data processing units (not shown) and volatile and non-volatile memories (not shown) for the execution of computer programs.
- the system 1 comprises a software system SW for implementing the method for generating 3D models according to the invention.
- the software system SW is hosted in the HD data storage device which typically includes one or more hard drives.
- the method of the invention is implemented in particular by the execution by the processor PROC of code instructions of the software system SW.
- the system 1 provides a 3D model generation service of the “SAAS” type which is made accessible to users through the IP network.
- An SP software module is included in the SW software system and is dedicated to the implementation of this service.
- this software module SP may include a user interface and / or a programming interface, called "API” (for "Application Programming Interface” in English), according to the embodiment of the invention.
- IT systems or devices UD such as computers, tablets and / or smartphones
- the users will be able to use the Internet browsers of their computer devices UD, or otherwise, for example, a software application exploiting 3D models for different purposes and which accesses the 3D model generation service of the system 1 through an interface " API ”, as noted above.
- the users are typically clients of the 3D model generation service provided by the system 1.
- the software system SW includes a convolutional neural network A1, known as the “CNN” network (for “Convolutional Neural Network” in English), authorizing the processing of multidimensional data and capable of in-depth training, or learning (known as “deep learning”). in English).
- the convolutional neural network A1 is preferably here of the so-called “supervised” type, but not exclusively.
- the convolutional neural network A1 could, for example, be developed using the Keras® library known to those skilled in the art as being an open source software library or “open source” in English.
- the convolutional neural network A1 provides an artificial intelligence function which is pre-trained to recognize an object represented in a 2D image supplied as input and to suggest one or more 3D models indexed in a database of 3D models, marked DB3D at the Fig. 2.
- the 3D model or models offered by the artificial intelligence function are closest to the object represented in the 2D image.
- the software system SW also includes an INS model instantiation and processing software module which cooperates with the convolutional neural network A1.
- the INS software module provides an adjustment function of the 3D model extracted from the database of 3D models to using the convolutional neural network A1, so as to generate and output a 3D model corresponding finely to the object represented by the 2D image.
- I2D image data is inputted by a user UD and typically includes 2D two-dimensional image data, strictly speaking, as well as metadata that describes the 2D image.
- the I2D image data is first processed by a certain number of convolutional layers CL which make it possible to extract, by successive filterings, characteristics present in the 2D image, such as a shape , texture, orientation and others.
- CC classification layers of the convolutional neural network A1 then exploit the characteristics extracted by the convolution layers CL, as indexing characteristics, to identify, by successive filtering, classes of close 3D models, corresponding to the image data I2D , and finally suggest a 3D model (or several models) typically in the form of a non-textured 3D model, marked N3D in Fig. 2, and CA characteristics of material, texture and others to add to it.
- the N3D model is the closest 3D model to the I2D image data identified by the convolutional neural network A1 in the DB3D model database.
- the convolutional neural network A1 identifies the characteristics of materials, textures and others to be given to the N3D model in order to obtain a 3D model, marked 03D in FIG. 2, which corresponds finely to the I2D image data.
- the characteristics of materials, textures and other identified CAs are added to the N3D model by means of 3D model processing software programs globally marked TM in Fig. 2.
- the database of DB3D models contains a textured 3D model identified by the Al network and which corresponds to the I2D image data, it is this 3D model which will be provided directly instead of the N3D model. aforementioned.
- the DB3D model database typically includes a library of textured 3D models and a library of untextured and unorganized 3D models. These libraries of 3D models in the DB3D model database are enriched over time.
- the DB3D model database is part of a DS dataset, which is used for training the convolutional neural network A1.
- the DS dataset essentially comprises the model database DB3D and 2D image data and is hosted here in the HD data storage device, together with the software library (not shown), for example the Keras® library, of the convolutional neural network A1.
- the N3D model is processed by the INS model instantiation and processing software module in order to produce the finished 3D model, marked 03D, which is supplied to the user UD by the system 1 in response to the I2D image data.
- the INS module comprises in particular the 3D model processing software TM software programs which process the N3D model in order to obtain the 03D model, based on the aforementioned characteristics CA identified by the convolutional neural network A1.
- TM software programs include in particular MAT, TX and TRE programs which respectively perform the functions of adding materials such as metallic, shiny, matt or other material, adding textures such as an image, and adding special effects like a transparency, a relief or others.
- the MAT, TX and TRE programs here are typically scripts in the Microsoft® C # programming language.
- the functionalities of a 3D creation software application such as for example the functionalities of a Unity3D® type software application, could also be used by the INS software module to execute a 3D model generation or a processing.
- the 03D models delivered by the INS model instantiation and processing software module can also be used by the system 1 so as to enrich the DS data set comprising the database of DB3D models.
- the method for generating 3D models according to the invention essentially comprises four major steps, S1 to S4, for implementing the operation and processing operations described above.
- Step S1 concerns the initial realization and enrichment of the DS dataset.
- Step S1 involves an initial assembly and organization of image data, to obtain a first DS data set.
- the DS data set is then reorganized to take account of its enrichment, in particular by adding 03D models delivered by the INS model instantiation and processing software module (see step S4).
- the DS dataset must include a rich assemblage of very diverse image data in order to enable efficient training of the convolutional neural network A1.
- the DS dataset obtained essentially comprises 2D image data and the database of A1.
- DB3D images mentioned above containing a library of textured 3D models and a library of un-textured and unorganized 3D models.
- Step S2 involves training the convolutional neural network A1.
- Step S2 includes substeps S20 and S21.
- the convolutional neural network A1 is configured for multiclassification of images based on several characteristics, namely, shape (designated SHAPE), material (designated MATERIAL), texture (designated TEXTURE) and various visual effects and / or effects that can be obtained by image processing, and the like.
- the convolutional neural network A1 is also configured to provide a choice function (designated F_SELECTION) for the N3D models and to identify the aforementioned CA characteristics, as indicated above with reference to Fig. 2, based on recognition probabilities obtained. .
- F_SELECTION a choice function for the N3D models and to identify the aforementioned CA characteristics, as indicated above with reference to Fig. 2, based on recognition probabilities obtained.
- the configuration of the convolutional neural network A1 can be modified according to the needs of the application, for example, by expanding the number of classes in the filtering levels, or layers, of the A1 network or by increasing the depth of the branches.
- the convolutional neural network A1 is first of all trained with the data set DS assembled initially, and subsequently to learn with the new image data having enriched the data set DS.
- the DS data set is divided into three disjoint sets, namely, a first set of so-called training image data, a second so-called validation image data set and a third so-called test image data set.
- the first set allows an iterative training algorithm of the A1 network to learn the characteristics of each class of images.
- the training algorithm performs a validation test with image data from the second set to determine a classification success rate.
- the training algorithm improves the success rate at each iteration by modifying weights associated with each neuron of the A1 network.
- the invention allows for continuous improvement of the artificial intelligence process.
- the pre-trained convolutional neural network A1 provides a prediction file, for example of CSV type (for “Comma-Separated Values” in English), containing a list of known 2D images. and various associated object classes.
- a 3D model generator uses the prediction file to create a corresponding 3D model.
- the 3D model is viewed by a system administrator 1 before validation and recording in the DB3D model database. If an error in the artificial intelligence process is noted by the administrator, it is corrected by modifying the prediction file which can then be reused for training the convolutional neural network A1.
- step S21 the training process is completed.
- An AMVIODEL neural model is then obtained, to which corresponds a weight matrix, which is recorded with its final success rate, called "accuracy" in English, in order to be used by the system 1.
- the classification of 2D images by the artificial intelligence process is then done by matrix operations from the pixel matrices of the 2D images and the weight matrix of the AMVIODEL neural model.
- the AMVIODEL neural model is of the multi-output type, several neural sub-models being associated with the different outputs.
- the training of the convolutional neural network A1 can be done by jointly training the various neural sub-models synchronously or by training them independently asynchronously.
- Step S3 concerns the exploitation of the neuronal model AMVIODEL.
- I2D image data is supplied as an input to the convolutional neural network A1.
- the neural model AMVIODEL provides the function F_SELECTION and authorizes the indexing of the 3D models, contained in the DB3D model database, closest to the I2D image data supplied as input.
- the closest 3D, non-textured (N3D) or textured (03D) model or models are identified, and if necessary the software programs TM for the aforementioned processing to be applied to the 3D models.
- a multiclassification of the incoming I2D images is carried out by the artificial intelligence process of the convolutional neural network A1 in order to determine suggestions of close 3D models and, optionally, of the processing operations to be applied to them in the process. using TM software programs for 3D model processing.
- the convolutional neural network A1 issues suggestions SG1 to SGn for an incoming image I2D, which suggestions arise from the multiclassification performed for this incoming image I2D.
- the SG1 suggestion arises from a classification of the incoming image I2D of an object, for example a glass bottle, into a main class of C_CYL objects and a secondary class of C_BEA objects.
- the main class C_CYL groups here objects having a cylindrical shape and includes the secondary class C_BEA which groups beer bottles.
- a 3D model, 03D2, associated with the secondary class C_BEA is proposed in this suggestion SG1.
- a PS1 probability that the 03D2 model is the closest model to the glass bottle of the incoming image I2D is attached to the SG1 suggestion.
- the SG2 suggestion arises from a classification of the incoming I2D image of the glass bottle into a main class of C_GEN objects and a secondary class of C_VAP objects.
- the main class C_GEN is a generic class comprising different types of objects and includes the secondary class C_VAP which groups cylindrical shaped vaporizers.
- a non-textured 3D model, N3D3, associated with the secondary class C_VAP, is proposed in this SG2 suggestion, as well as the application of a Mat3 material to this N3D3 model.
- a PS2 probability that the model N3D3 with the material Mat3 is the closest model to the glass bottle of the incoming image I2D is attached to the suggestion SG2.
- the SG3 suggestion arises from a classification of the incoming image I2D of the glass bottle into the main object class C_CYL and the secondary object class C_BEA mentioned above.
- the suggestion with the highest probability is the one that is prioritized by the artificial intelligence process. It will be noted that the aforementioned probabilities could be replaced by weightings, percentages, success rates or the like according to the embodiments of the invention.
- Step S4 concerns the processing of the N3D model by the INS model instantiation and processing software module in order to obtain the 03D model which finely corresponds to the I2D image data.
- the 03D models obtained during this step S4 can be used to enrich the DS data set, as indicated above in relation to step S1.
- Various tasks are performed in step S4 by the software module INS. More generally, scripts for modeling and processing 3D models typically in the Microsoft® C # programming language can be executed here, as well as the functionalities of a Unity3D® type software application.
- the 03D model obtained is recorded preferentially, but not exclusively, in the Collada® format, known as “. dae ”.
- a 3D visualization interface (not shown) is included in the INS software module for 3D models. This 3D visualization interface is typically developed in the C # programming language, in a Unity3D® type environment.
- the convolutional neural network A1 emits suggestions SG (n-1), SGn and SG (n + 1) for an incoming image I2D, respectively with probabilities PS (n-1 ), PSn and PS (n + 1).
- Suggestion SG (n-1) provides an untextured 3D model, N3Dp, with a standard MatS material to apply.
- Suggestions SG (n) and SG (n + 1) both provide an untextured 3D model, N3Dq, with a matte material MatM and a material with reflection and transparency MatRT respectively.
- the program for adding MAT materials includes scripts St (m-1), Stm and St (m + 1) dedicated respectively to adding the materials MatS, MatM and MatRT.
- the three models 03Dp, 03Dq1 and 03Dq2 are presented to the user USER through a 3D visualization interface, VIS.
- the process retains as the 03D model the one to which the highest probability is attached. Considering, for example, that the highest probability is PS (n-1), the 03Dp model is retained.
- the input 2D image is the l2Da image of a product which is a glass bottle, for example, a cider bottle.
- the artificial intelligence process identifies a class of products CL3 which corresponds to the glass bottle of image l2Da, among several classes of products CL1 to CL4.
- the artificial intelligence process recognizes the glass bottle in image l2Da as belonging to product classes CL1 through CL4 with probabilities P1 through P4, respectively.
- the probability P3 being the highest among the probabilities P1 to P4, it is the product class CL3 which is chosen as the membership class of the glass bottle in image l2Da. It is here essentially the shape of the glass bottle of the l2Da image, of cylindrical type, which will have enabled the artificial intelligence process to carry out this first selection.
- the artificial intelligence process seeks to determine which product of the class CL3 is closest to the glass bottle of the l2Da image. For this, the artificial intelligence process here classifies products of class CL3 into two classes of products GP1 and GP2.
- the artificial intelligence process recognizes the glass bottle in image l2Da as belonging to product classes GP1 and GP2 with probabilities P10 and P11, respectively. Since the probability P10 is higher than the probability P11, the product class GP1 is chosen as the membership class of the glass bottle in image l2Da. It is here essentially the glass material of the bottle in the l2Da image that allowed the artificial intelligence process to make this second selection.
- Product class GP1 includes various glass bottles.
- the artificial intelligence process here classifies the glass bottles of the GP1 class into three product classes GF1 to GF3, mainly based on the shape of the glass bottles. Three different bottle shapes FP1 to FP3 correspond to the product classes GF1 to GF3, respectively.
- the artificial intelligence process recognizes the glass bottle of image l2Da as having the probabilities P17 through P19 of having the bottle shapes FP1 through FP3, respectively.
- the probability P17 being the higher of the probabilities P17 and P19, it is the bottle shape FP1 which is chosen as being the shape of the glass bottle of the image l2Da.
- N3Da non-textured 3D model, marked N3Da in Fig. 3, which is stored in the database of models BD3D.
- the N3Da model is indexed by the artificial intelligence process to the l2Da image and is downloaded.
- the INS model instantiation and processing software module resizes the N3Da model to match the product of the l2Da image and applies to it the glass material recognized by the artificial intelligence process and a texture extracted from the l2Da image .
- the INS module incorporates a software application of the aforementioned Unity3D® type
- the N3Da model is loaded into a scene of this software application in order to apply the material and the texture to it.
- - R is a regularization parameter.
- the nodes are the product classes CL1 to CL4, GP1, GP2, and GF1 to GF3.
- the application of the choice function F_SELECTION for the calculation of probabilities, with R 0) is detailed in block B2 of Fig. 7.
- the probabilities for the three branches CL3 / GP1 / GF1, CL3 / GP1 / GF2 and CL3 / GP1 / GF3 are given by F_SELECTION (GF1), F_SELECTION (GF2) and F_SELECTION (GF3), respectively.
- the probability F_SELECTION (GF1) P3 * P10 * P17 of the branch CL3 / GP1 / GF1 is considered to be greater than the probabilities of other branches.
- the artificial intelligence process therefore chooses the class GF1 and the 3D model, N3Da, as corresponding to the glass bottle in the image l2Da.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- Architecture (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
- Processing Or Creating Images (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FR1914311A FR3104786B1 (fr) | 2019-12-12 | 2019-12-12 | Procédé et système de génération de modèles numériques 3d |
| PCT/FR2020/052416 WO2021116635A1 (fr) | 2019-12-12 | 2020-12-14 | Procede et systeme de generation de modeles numeriques 3d |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4073765A1 true EP4073765A1 (fr) | 2022-10-19 |
Family
ID=70918493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP20845411.6A Withdrawn EP4073765A1 (fr) | 2019-12-12 | 2020-12-14 | Procédé et système de génération de modèles numériques 3d |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US20230019232A1 (fr) |
| EP (1) | EP4073765A1 (fr) |
| CA (1) | CA3160129A1 (fr) |
| FR (1) | FR3104786B1 (fr) |
| WO (1) | WO2021116635A1 (fr) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12505583B2 (en) * | 2020-06-02 | 2025-12-23 | Nvidia Corporation | Techniques to process layers of a three-dimensional image using one or more neural networks |
| JP6799883B1 (ja) * | 2020-07-27 | 2020-12-16 | 株式会社Vrc | サーバ及び情報処理方法 |
| CN116824034A (zh) * | 2022-12-23 | 2023-09-29 | 慧之安信息技术股份有限公司 | 一种智慧园区数化仿真方法及系统 |
| KR102702240B1 (ko) * | 2023-02-13 | 2024-09-04 | 주식회사 메이코더스 | 웹 기반 환경에서의 2차원 이미지를 이용한 용기 제조용 3차원 모델링 방법, 장치 및 컴퓨터-판독가능 기록매체 |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5436559B2 (ja) * | 2008-09-02 | 2014-03-05 | エコール・ポリテクニーク・フェデラル・ドゥ・ローザンヌ(エーペーエフエル) | ポータブル・デバイス上での画像アノテーション |
| US8570343B2 (en) | 2010-04-20 | 2013-10-29 | Dassault Systemes | Automatic generation of 3D models from packaged goods product images |
| US11263823B2 (en) * | 2012-02-24 | 2022-03-01 | Matterport, Inc. | Employing three-dimensional (3D) data predicted from two-dimensional (2D) images using neural networks for 3D modeling applications and other applications |
| EP3179407B1 (fr) * | 2015-12-07 | 2022-07-13 | Dassault Systèmes | Reconnaissance d'un objet modélisé 3d à partir d'une image 2d |
| US20180114363A1 (en) * | 2016-10-25 | 2018-04-26 | Microsoft Technology Licensing, Llc | Augmented scanning of 3d models |
| US10769411B2 (en) * | 2017-11-15 | 2020-09-08 | Qualcomm Technologies, Inc. | Pose estimation and model retrieval for objects in images |
| US11367222B2 (en) * | 2018-04-20 | 2022-06-21 | Hewlett-Packard Development Company, L.P. | Three-dimensional shape classification and retrieval using convolutional neural networks and majority vote |
| US10810469B2 (en) * | 2018-05-09 | 2020-10-20 | Adobe Inc. | Extracting material properties from a single image |
| US11321769B2 (en) * | 2018-11-14 | 2022-05-03 | Beijing Jingdong Shangke Information Technology Co., Ltd. | System and method for automatically generating three-dimensional virtual garment model using product description |
| US11507781B2 (en) * | 2018-12-17 | 2022-11-22 | Bodygram, Inc. | Methods and systems for automatic generation of massive training data sets from 3D models for training deep learning networks |
| US11361505B2 (en) * | 2019-06-06 | 2022-06-14 | Qualcomm Technologies, Inc. | Model retrieval for objects in images using field descriptors |
-
2019
- 2019-12-12 FR FR1914311A patent/FR3104786B1/fr active Active
-
2020
- 2020-12-14 WO PCT/FR2020/052416 patent/WO2021116635A1/fr not_active Ceased
- 2020-12-14 US US17/783,383 patent/US20230019232A1/en not_active Abandoned
- 2020-12-14 CA CA3160129A patent/CA3160129A1/fr active Pending
- 2020-12-14 EP EP20845411.6A patent/EP4073765A1/fr not_active Withdrawn
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021116635A1 (fr) | 2021-06-17 |
| CA3160129A1 (fr) | 2021-06-17 |
| FR3104786A1 (fr) | 2021-06-18 |
| FR3104786B1 (fr) | 2022-01-21 |
| US20230019232A1 (en) | 2023-01-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12450870B2 (en) | Classifying image styles of images based on procedural style embeddings | |
| EP4073765A1 (fr) | Procédé et système de génération de modèles numériques 3d | |
| US12547435B2 (en) | Personalized modification of audio and visual components of a virtual agent of a user interface | |
| EP3686848A1 (fr) | Synthèse d'images sémantiques pour générer des images sensiblement photoréalistes à l'aide de réseaux neuronaux | |
| CN108122264B (zh) | 促进草图到绘画变换 | |
| US20180293712A1 (en) | Denoising monte carlo renderings using generative adversarial neural networks | |
| US12198231B2 (en) | Generating differentiable procedural materials | |
| CN118710781A (zh) | 利用端到端机器学习模型的面部表情和姿势转移 | |
| US9940551B1 (en) | Image generation using neural networks | |
| US12243095B2 (en) | Techniques for generating a three dimensional model for a listing | |
| EP3639190A1 (fr) | Procede d'apprentissage de descripteurs pour la detection et la localisation d'objets dans une video | |
| JP7457211B2 (ja) | サードパーティ資産を用いて拡張現実体験を促進するためのコンピューティングプラットフォーム | |
| CN117853612A (zh) | 利用人类修复模型生成经修改的数字图像 | |
| US20220309345A1 (en) | Machine Learning Model Based Embedding for Adaptable Content Evaluation | |
| WO2025199044A1 (fr) | Systèmes et procédés de construction assistée par l'ia de modèles tridimensionnels | |
| US20250111575A1 (en) | System and method for the automated creation of shaders in 3d graphics design | |
| US12361607B2 (en) | Generating digital images in new contexts while preserving color and composition using diffusion neural networks | |
| US11972534B2 (en) | Modifying materials of three-dimensional digital scenes utilizing a visual neural network | |
| WO2023091325A1 (fr) | Rendu non photo-réaliste en temps réel | |
| US12260484B2 (en) | Real-time non-photo-realistic rendering | |
| EP4664349A1 (fr) | Récipients génératifs | |
| EP4589537A1 (fr) | Procédé de génération de données, programme d'ordinateur et dispositif de calcul associés | |
| FR3133261A1 (fr) | Essai de style avec Des GAN inversÉs | |
| FR3159460A3 (fr) | Procédés, appareil et systèmes incorporant des gan minimisant l’enchevêtrement d'attributs à l’aide d’espaces gan autocorrigés | |
| GB2700348A (en) | Real-time text-based disentangled real image editing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20220708 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) | ||
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
| 17Q | First examination report despatched |
Effective date: 20250130 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN |
|
| 18D | Application deemed to be withdrawn |
Effective date: 20250531 |