Research map / Vision-language and multimodal / Cross-modal learning
Cross-view Semantic Alignment for Livestreaming Product Recognition
Wenjie Yang, Yiyi Chen, Yan Li, Yanhua Cheng, Xudong Liu, Quan Chen, Han Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add pairwise matching decoder (patch-level matching) to instance-level contrastive learning (ablation) | RICE · LPR4M test | R1: 27.1 -> 29.4 | Improved |
| add Patch Feature Reconstruction loss (ablation) | RICE · LPR4M test | R1: 29.4 -> 30.3 | Improved |
| add intended product detection (box input instead of patches) (ablation) | RICE · LPR4M test | R1: 30.3 -> 31.3 | Improved |
| add text modality (ASR and title) to similarity (ablation) | RICE · LPR4M test | R1: 31.3 -> 33.0 | Improved |
| add text modality to plain ICL (ablation) | ICL baseline · LPR4M test | R1: 27.1 -> 28.5 | Improved |
| use detected product box input instead of patches (ablation) | ICL · LPR4M small-scale subset | R1: 23.9 -> 30.1 | Improved |
| use detected product box input instead of patches (ablation) | RICE · LPR4M abundant-products subset | R1: 21.0 -> 17.6 | Hurt |
| use RICE over strongest video-understanding baseline SwinB | livestreaming product recognition · LPR4M test | R1: 29.1 -> 33.0 | Improved |
| use RICE (frame feature averaging) instead of SwinB (3D shifted windows) on occlusion-heavy categories | RICE · LPR4M per-category, e.g. Suitcase&bag | R1 | Mixed |
Closest papers by meaning
- CovMatch: Cross-Covariance Guided Multimodal Dataset Distillation with Trainable Text Encoder (NeurIPS 2025 · similarity 0.85)
- Dynamic Adapter with Semantics Disentangling for Cross-lingual Cross-modal Retrieval (AAAI 2025 · similarity 0.84)
- Beyond Global Similarity: Multi-Conditional Retrieval for Fine-Grained Cross-Modal Understanding (CVPR 2026 · similarity 0.84)
- Global-Semantic Alignment Distillation for Partial Multi-view Classification (AAAI 2025 · similarity 0.83)
- Linguistic-Aware Patch Slimming Framework for Fine-grained Cross-Modal Alignment (CVPR 2024 · similarity 0.83)
- TOPA: Extending Large Language Models for Video Understanding via Text-Only Pre-Alignment (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
