AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Semantic-Augmented Image Clustering via Adaptive Multi-Modal Collaboration

Xiaohan Zhang, Chao Zhang, Deng Xu, Hong YU, Chunlin Chen, Huaxiong Li

AAAI 2026 · Vision-language and multimodal › Cross-modal learning · approach: Unimodal-to-multimodal representation transfer · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add LBL balance loss to LRC+LCL (ablation)SAC · CIFAR-20clustering ACC (%): 56.3 -> 67.4Improved
add LBL balance loss to LRC+LCL (ablation)SAC · ImageNet-Dogsclustering ACC (%): 72.8 -> 86.5Improved
add LCL alignment loss to LRC+LBL (ablation)SAC · CIFAR-20clustering ACC (%): 66.2 -> 67.4Improved
add LCL alignment loss to LRC+LBL (ablation)SAC · ImageNet-Dogsclustering ACC (%): 84.4 -> 86.5Improved
add LRC loss to LBL+LCL (ablation)SAC · CIFAR-20clustering ACC (%): 56.2 -> 67.4Improved
add LRC loss to LBL+LCL (ablation)SAC · ImageNet-Dogsclustering ACC (%): 48.7 -> 86.5Improved
add adaptive neighborhood uncertainty weights wIN and wCN (ablation)SAC · CIFAR-20clustering ACC (%): 65.7 -> 67.4Improved
add adaptive weights wIN and wCN (ablation)SAC · ImageNet-Dogsclustering ACC (%): 84.9 -> 86.5Improved
add intra-modal contrastive loss to inter-modal only (ablation)SAC · ImageNet-Dogsclustering ACC (%): 78.1 -> 86.5Improved
add inter-modal contrastive loss to intra-modal only (ablation)SAC · CIFAR-20clustering ACC (%): 66.4 -> 67.4Improved
use BLIP-2 instead of WordNet text generator (ablation)SAC · CIFAR-20clustering ACC (%): 55.7 -> 67.4Improved
use BLIP-2 instead of BLIP text generator (ablation)SAC · CIFAR-20clustering ACC (%): 55.8 -> 67.4Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy