Search papers, labs, and topics across Lattice.
Tencent Inc
4
0
7
Visual reranking and active rejection in MMAgent-R$^2$ significantly boost retrieval accuracy in challenging KB-VQA tasks, outperforming traditional methods.
Shifting from bounding boxes to pixel-level segmentation in MLLMs leads to significant gains in visual reasoning accuracy and segmentation performance.
Forget brute-force scaling: REVERSE shows that teaching an agent *how* to search and verify evidence lets a smaller model beat giants at image geo-localization.
Current video understanding models struggle with long-horizon robustness and non-speech audio, as revealed by the new OmniPro benchmark designed for comprehensive omni-modal proactive evaluation.
