实验室2篇论文被2026 ECCV接收。该会议是计算机视觉领域三大顶会之一，每年举办一届。该届会议将于2026年9月8-12日于Malmö, Sweden举行。

题目: TARS: MinMax Token-Adaptive Preference Strategy for Hallucination Reduction in MLLMs

作者：Kejia Zhang, Keda TAO, Zhiming Luo, Chang Liu, Jiasheng Tang, Huan Wang

摘要：Multimodal large language models (MLLMs) are prone to hallucinations, generating plausible but visually ungrounded outputs, partly because direct preference optimization (DPO) overfits to superficial linguistic cues under static preference supervision. We propose TARS, a token-adaptive preference strategy that reformulates DPO as a principled min-max optimization problem. The inner maximization selectively perturbs visual-agnostic tokens to induce worst-case distributional shifts, while the outer minimization enforces alignment with causal visual signals rather than surface-level patterns. A novel spectral alignment loss further regularizes hidden representations in the frequency domain via the Fast Fourier Transform (FFT), preserving global semantic structure without rigid token-level correspondence. We evaluate TARS across multiple hallucination benchmarks. Using only 4.8k preference samples without expert feedback, TARS reduces hallucination rates from 26.4% to 13.2% and cognition scores from 2.5 to 0.4, outperforming standard DPO by a large margin. Notably, TARS surpasses LLM-based data augmentation trained on 28.8k samples (Hal-Rate: 16.0% vs. 13.2%), demonstrating that reshaping the optimization landscape via adversarial token perturbation is fundamentally more effective than scaling training data. TARS further narrows the gap with GPT-4o on key metrics.

题目: Rectified Embedding Flow Learning for Multi-view Geo-localization

作者：Hao Ruan, Jinliang Lin, Yingxin Lai, Zhiming Luo, Shaozi Li, Yu Zang, Cheng Wang

摘要：Aerial geo-localization is increasingly essential for large-scale spatial retrieval. To overcome the limitations of single-retrieval systems caused by modality-specific uncertainty in open environments, this paper introduces the unified Aerial Multi-view Geo-localization (AMGL) task. However, when applying universal multimodal retrieval paradigms to this task, the semantic bias introduced solely by textual instructions fails to resolve and reconstruct the differentiated distribution structures necessary for cross-view alignment. Consequently, this paper proposes the Rectified Embedding Flow Learning (REFL) framework, formulating cross-domain alignment as a directed conditional distribution transport process. Initially, REFL employs velocity-prior flow learning to fit continuous ordinary differential equation trajectories, deriving transformation priors that map single-view distributions to a latent shared manifold. Subsequently, a trajectory-guided embedding rectification mechanism continuously transports query features to the target view distribution, explicitly compensating for distribution shifts. Extensive evaluations on the Aerial MVGL benchmark demonstrate that REFL achieves state-of-the-art performance, yielding an average R@1 of 44.83% and R@10 of 65.85%.