6 papers
SpaceCache: Intermediate Feature Reuse for Efficient Semantic Segmentation on Point Cloud Streams
Dongheon Yang, Beomseok Kang, Jae-Joon Kim
Point-cloud streams are central to autonomous driving and robotics, yet existing segmentation networks process every frame from scratch, although consecutive frames largely overlap…
Accelerating VLA Inference with Perturbation Informed Execution
Jehun Lee, Hajun Lee, Jimin Ryu +4
Action-chunking vision-language-action (VLA) models predict sequences of future robot actions. Under fixed-horizon execution, the robot executes a fixedlength prefix of each predic…
Rotation-Aligned Key Channel Pruning for Efficient Vision-Language Model Inference
Beomseok Kang, Dongwon Jo, Jiwon Song +2
Vision-Language Models suffer severe KV cache pressure at inference, as a single image often encodes into thousands of tokens. Most existing methods exploit token sparsity through…
CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection
Jiwon Song, Dongwon Jo, Beomseok Kang +1
Chunked prefill has become a widely adopted serving strategy for long-context large language models, but efficient attention computation in this regime remains challenging. Existin…
Efficient Domain-Adaptive Multi-Task Dense Prediction with Vision Foundation Models
Beomseok Kang, Niluthpol Chowdhury Mithun, Mikhail Sizintsev +2
Multi-task dense prediction, which aims to jointly solve tasks like semantic segmentation and depth estimation, is crucial for robotics applications but suffers from domain shift w…
DUDA: Distilled Unsupervised Domain Adaptation for Lightweight Semantic Segmentation
Beomseok Kang, Niluthpol Chowdhury Mithun, Abhinav Rajvanshi +2
Unsupervised Domain Adaptation (UDA) is essential for enabling semantic segmentation in new domains without requiring costly pixel-wise annotations. State-of-the-art (SOTA) UDA met…