Ultra-high-resolution (UHR) remote sensing visual question answering (VQA) requires locating fine-grained evidence in large scenes under limited visual input budgets. Active observation addresses this challenge through selective exploration, but repeated generative judgments can make region search costly. Its decisions...
Peng-Yu Dai, Yao Zhang, Jian Song et al.· 0 citations
This work proposes TRIP-Shapley, a propagation-aware Shapley-inspired surrogate for contribution attribution in P2PFL with a model-free coordinator used solely for auditing, and introduces lightweight mechanisms to detect and mitigate misreporting attacks.
Honoka Anada, Tatsuya Kaneko, Ya-Fei Ou et al.· 0 citations
Balancing performance trade-offs on long-tailed data distributions remains a long-standing challenge in visual recognition. Existing methods mainly improve tail classes through re-balancing, representation learning, or data augmentation, but the underlying cause of tail class degradation is still insufficiently explore...
Sheng-Han Chen, Yi-Ming Liu, Zhi-Peng Deng et al.· 0 citations
This work proposes BeatRhythm-TTA, an ECG-tailored TTA framework that explicitly accounts for ECG's noisy observations and structured beat-rhythm semantics under domain shift, and introduces a Signal Quality Index-gated adaptation scheme that selectively filters out low-quality signals to prevent harmful updates.
Wen-Han Jiang, Zhi-Peng Deng, Jia-Le Zhou et al.· International Conference on...· 1 citation
Multimodal large language models (MLLMs) must balance local detail against scene context when interpreting ultra-high-resolution (UHR) remote sensing (RS) imagery within a limited visual-input budget. Existing selection-based methods either prune tokens and select patches through relevance scoring, or crop actively thr...
Yao Zhang, Peng-Yu Dai, Wei Guo et al.· arXiv.org· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.