System and method for improved generation of avatars for virtual try-on of garments

A method for improved generation of 3D avatars for virtual try-on of garments, the method is implemented by a processor executing program instructions stored in a memory, the method comprising:

• receiving inputs from a first user type, via a first input unit, for generating one or more garment types in a graphical format, wherein the inputs for generating the one or more garment types comprises a fully stitched garment image, a 2D flat image of the garment or a 3D sketch image of the garment;
• generating a 3D avatar of a second user type by capturing a first input type and a second input type respectively received via a second input unit, wherein the first input type comprises measurements of body specifications of the second user type and the second input type comprises body images of the second user type captured from different angles via a camera present on the second input unit in a 2D format and received by employing a body measurement Application Programming Interface (API),
• wherein the 3D avatar of the second user type is generated by parallelly implementing a combination of deep learning techniques including Convolutional Neural Network (CNN) technique, a Graph Neural Network technique and a Capsule Neural Network technique on the second input type, selecting the technique that provides a high accuracy for computing one or more body measurements of the second user type to obtain a segmented output image, implementing a combination of classification techniques on the segmented output image including Bayesian, k-Nearest Neighbors (k-NN), Support Vector Machine (SVM), decision trees, adaptive boosting, random forest, Multilayer Perceptron (MLP) and expectation-maximization classifier, selecting the classification technique that provides a high accuracy in computation of body contours of the second user type for computing an external body contour image of the second user type, and extracting one or more key points from the computed external body contour image of the second user type, and wherein the Graph Neural Network technique is implemented to process the extracted one or more key points for determining the body measurements of the second user type;
• and
• rendering the generated garment types on the generated 3D avatar of the second user type for carrying out a virtual try-on operation.

A computer program product comprising a non-transitory computer-readable medium having computer program code stored thereon, the computer-readable program code comprising instructions that, when executed by a processor, causes the processor to:

• receive inputs from a first user type, via a first input unit, for generating one or more garment types in a graphical format, wherein the inputs for generating the one or more garment types comprises a fully stitched garment image, a 2D flat image of the garment or a 3D sketch image of the garment;
• generate a 3D avatar of a second user type by capturing a first input type and a second input type respectively received via a second input unit, wherein the first input type comprises measurements of body specifications of the second user type and the second input type comprises body images of the second user type captured from different angles via a camera present on the second input unit in a 2D format and received based on employing a body measurement Application Programming Interface (API),
• wherein the 3D avatar of the second user type is generated by parallelly implementing a combination of deep learning techniques including Convolutional Neural Network (CNN) technique, a Graph Neural Network technique and a Capsule Neural Network technique on the second input type, selecting the technique that provides a high accuracy for computing one or more body measurements of the second user type to obtain a segmented output image, implement a combination of classification techniques on the segmented output image including Bayesian, k-Nearest Neighbors (k-NN), Support Vector Machine (SVM), decision trees, adaptive boosting, random forest, Multilayer Perceptron (MLP) and expectation-maximization classifier, selecting the classification technique that provides a high accuracy in computation of body contours of the second user type for computing an external body contour image of the second user type, and extracting one or more key points from the computed external body contour image of the second user type, and wherein the Graph Neural Network technique is implemented to process the extracted one or more key points for determining the body measurements of the second user type;
• and
• render the generated garment types on the generated 3D avatar of the second user type for carrying out a virtual try-on operation.