Customized AR Video Overlay for Multi-Product Virtual Try-On
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AR technologies limit user interaction and sharing of AR videos, requiring tedious manual processes for trying on multiple products and lacking mechanisms for sophisticated user feedback and transaction completion.
Innovation Solution
A system that automatically generates a customized AR video by overlaying virtual models on real-world video segments, allowing interactive editing and sharing, and enabling transactions based on user feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a user wants to virtually try on multiple products using existing AR technologies, then the user can view each product in AR, but the user must go through tedious interactions for each product and consume added computing resources
Solution Approach 1:
The system pre-generates multiple AR video segments with different virtual models overlaid on the user's video footage before the user needs to view them. This preliminary generation eliminates the need for users to perform repeated capture and rendering operations for each product, significantly reducing interaction time and computational resource consumption during the product comparison phase.
Solution Approach 2:
The system combines multiple individual AR try-on sessions into a single continuous AR video that displays multiple products sequentially. By merging the video footage and virtual model overlays into one integrated output, the system allows users to compare multiple products in a single viewing session rather than performing separate interactions for each product.
2Reliability
If existing AR technologies allow users to capture static video of AR experience, then the user can save the video, but the AR content is not editable as AR at a later time
Solution Approach 1:
The system segments the AR video into distinct components: the base video footage and multiple virtual model overlays. Each virtual model is stored as a separate layer or segment that can be independently accessed, selected, and modified. This segmentation enables the AR content to be edited at a later time while maintaining the reliability of the saved video file.
3Adaptability or versatility
If users want to solicit feedback from other users on product try-on experience, then the user can share the video, but the feedback mechanism is very basic
Solution Approach 1:
The system incorporates a structured feedback mechanism where recipients of shared AR videos can provide specific feedback on different products tried on. The feedback system allows users to rate, comment on, and compare multiple products within the same video, transforming basic feedback into sophisticated product evaluation data that helps the original user make informed purchasing decisions.
Data Source
AI summary
Methods and systems are disclosed for generating an augmented reality (AR) video. A set of products is obtained, where each product is associated with a respective virtual model and a respective object class. An AR video segment is generated for each product in the set of products. In a real-world video segment, a real-world object belonging to a relevant object class that is relevant to the object class of the given product is detected. A render of the virtual model associated with the given product is overlaid in the real-world video segment to obtain the AR video segment. The render of the virtual model is overlaid relative to the detected real-world object belonging to the relevant object class. A continuous AR video is generated from the AR video segments and outputted to be viewable by a user device.