3 citations · 8 across the 9 of their papers we have counts for
9 papers
DV-SFT: Direct Vision Supervision for Fine-Grained Visual Understanding
Jianfei Zhao, Feng Zhang, Xin Sun +3
Multimodal large language models are typically trained end-to-end to predict ground-truth answers, yet supervision signals are applied exclusively to text tokens. Visual tokens, th…
Attention Sinks and Outliers in Attention Residuals
Haozheng Luo, Haoran Dai, Shaoyang Zhang +10
We propose OASIS, an outlier- and sink-aware method that stabilizes dual-normalized attention-residual architectures through explicit null routing and token-to-depth null coupling.…
Benchmarking the Robustness of LiDAR-Camera Fusion for 3D Object Detection
Kaicheng Yu, Tang Tao, Hongwei Xie +10
There are two critical sensors for 3D perception in autonomous driving, the camera and the LiDAR. The camera provides rich semantic information such as color, texture, and the LiDA…
Knowledge Distillation via the Target-aware Transformer
Sihao Lin, Hongwei Xie, Bing Wang +4
Knowledge distillation becomes a de facto standard to improve the performance of small neural networks. Most of the previous works propose to regress the representational features…
Exploring Inter-Channel Correlation for Diversity-preserved KnowledgeDistillation
Li Liu, Qingle Huang, Sihao Lin +4
Knowledge Distillation has shown very promising abil-ity in transferring learned representation from the largermodel (teacher) to the smaller one (student).Despitemany efforts, pri…
Contrastive Instruction-Trajectory Learning for Vision-Language Navigation
Xiwen Liang, Fengda Zhu, Yi Zhu +3
The vision-language navigation (VLN) task requires an agent to reach a target with the guidance of natural language instruction. Previous works learn to navigate step-by-step follo…
DS-Net++: Dynamic Weight Slicing for Efficient Inference in CNNs and Transformers
Changlin Li, Guangrun Wang, Bing Wang +3
Dynamic networks have shown their promising capability in reducing theoretical computation complexity by adapting their architectures to the input during inference. However, their…
Dynamic Slimmable Network
Changlin Li, Guangrun Wang, Bing Wang +3
Current dynamic networks and dynamic pruning methods have shown their promising capability in reducing theoretical computation complexity. However, dynamic sparse patterns on convo…
BossNAS: Exploring Hybrid CNN-transformers with Block-wisely Self-supervised Neural Architecture Search
Changlin Li, Tao Tang, Guangrun Wang +4
A myriad of recent breakthroughs in hand-crafted neural architectures for visual recognition have highlighted the urgent need to explore hybrid architectures consisting of diversif…