Huayu Mai

Attention Contrastive Decoding: Preserving Coherence While Mitigating Hallucinations in Large Vision-Language Models

Towards Unsupervised Domain Bridging via Image Degradation in Semantic Segmentation

BeyondMix: Leveraging Structural Priors and Long-Range Dependencies for Domain-Invariant LiDAR Segmentation

Towards Unbiased Learning in Semi-Supervised Semantic Segmentation

Beyond Confidence: Exploiting Homogeneous Pattern for Semi-Supervised Semantic Segmentation