A commodity image evaluation method and device, a computing device and a readable storage medium
By classifying product images based on click-through rate and using an image comparison model to evaluate them, the problem of subjective evaluation in existing technologies is solved, and a more accurate assessment of product image quality is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ALIBABA GROUP HOLDING LTD
- Filing Date
- 2020-03-04
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for evaluating product images lack objectivity, relying primarily on aesthetic assessments which fail to accurately reflect product appeal and are not objective enough.
By identifying product categories, dividing quality levels based on click-through rates, obtaining reference images, and using image comparison models for evaluation, combined with Siamese neural networks and feature joint models for comparison, the objectivity and accuracy of the evaluation are improved.
It enables objective product image evaluation based on click-through rate, avoiding interference from inherent image features across different categories and improving the accuracy and consistency of the evaluation.
Smart Images

Figure CN113362268B_ABST