Research map / Label-efficient and robust learning / Recognition and long-tail
UniFast-HGR: Scalable and Efficient Maximal Correlation for Multimodal Models
Hongkang Zhang, Shao-Lun Huang, Yanlong Wang, Ercan Engin KURUOGLU
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 cosine reformulation to Soft-HGR baseline (ablation) | UniFast-HGR · Berlin | overall accuracy (%): 65.80 -> 75.23 | Improved |
| add cosine reformulation to Soft-HGR baseline (ablation) | UniFast-HGR · ImageNet-1K | top-1 accuracy (%): 76.3 -> 77.8 | Improved |
| add variance constraints (ablation) | UniFast-HGR · Berlin | overall accuracy (%): 75.23 -> 78.65 | Improved |
| add variance constraints (ablation) | UniFast-HGR · ImageNet-1K | top-1 accuracy (%): 77.8 -> 79.2 | Improved |
| add Gram-space structure-level term with diagonal retained (ablation) | UniFast-HGR · Berlin | overall accuracy (%): 78.65 -> 80.62 | Improved |
| add Trivial Spectrum Suppression (diagonal removal) (ablation) | UniFast-HGR · Berlin | overall accuracy (%): 80.62 -> 80.75 | Improved |
| add Trivial Spectrum Suppression (diagonal removal) (ablation) | UniFast-HGR · ImageNet-1K | top-1 accuracy (%): 79.8 -> 80.1 | Improved |
| use OptFast random-projection approximation instead of full Gram (ablation) | UniFast-HGR · ImageNet-1K | top-1 accuracy (%): 80.1 -> 79.6 | Mixed |
| reduce ImageNet local batch size from 256 to 16 (Soft-HGR) (ablation) | Soft-HGR · ImageNet-1K | top-1 accuracy (%): 76.3 -> 72.1 | Hurt |
| reduce ImageNet local batch size from 256 to 16 (UniFast-HGR with TSS) | UniFast-HGR · ImageNet-1K | top-1 accuracy (%): 80.1 -> 78.2 | Hurt |
| 80% single-class imbalance in batches | UniFast-HGR vs Soft-HGR · Berlin | overall accuracy drop (points): 6.60 vs 1.85 | Mixed |
| single-label homogeneous batches | UniFast-HGR vs Soft-HGR · Berlin | overall accuracy drop (points): 13.40 vs 3.55 | Mixed |
Closest papers by meaning
- An Efficient Approach to Informative Feature Extraction from Multimodal Data (AAAI 2019 · similarity 0.88)
- Multimodal Classification via Total Correlation Maximization (ICLR 2026 · similarity 0.86)
- BALM: A Model-Agnostic Framework for Balanced Multimodal Learning under Imbalanced Missing Rates (CVPR 2026 · similarity 0.85)
- TiCAL:Typicality-Based Consistency-Aware Learning for Multimodal Emotion Recognition (AAAI 2026 · similarity 0.85)
- Multi-to-Single: Reducing Multimodal Dependency in Emotion Recognition Through Contrastive Learning (AAAI 2025 · similarity 0.84)
- A Multi-Focus-Driven Multi-Branch Network for Robust Multimodal Sentiment Analysis (AAAI 2025 · 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.
