13 papers
ReSS: Residual-Restoring Sparse Attention for 3D Vision Transformers
Yongsung Kim, Jaehoon Lee, Minjun Park +3
3D vision transformers such as VGGT predict camera poses and scene geometry from multi-view images in a single forward pass, but their global attention over all concatenated view t…
FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance
Jaihyun Lew, Mingi Jung, Minjun Park +2
Reference-based image quality assessment (IQA) metrics aim to reflect how humans perceive the perceptual distance between a pair of images. To learn how the human visual system (HV…
NaturalFlow: Reducing Disruptive Pauses for Natural Speech Flow in Simultaneous Speech-to-Speech Translation
Dongwook Lee, Youngho Cho, Sangkwon Park +2
Simultaneous speech-to-speech translation aims to enable near-real-time communication by minimizing latency, offering a compelling, real-time alternative to the high latency of con…
Rollout-Level Advantage-Prioritized Experience Replay for GRPO
Gyeongtae Yoo, Sanghyeok Park, Soohyuk Jang +2
Reinforcement learning from verifiable rewards with GRPO is a standard approach for post-training reasoning LLMs. It remains sample inefficient. Each rollout is used for a single g…
HeSS: Head Sensitivity Score for Sparsity Redistribution in VGGT
Yongsung Kim, Wooseok Song, Jaihyun Lew +3
Visual Geometry Grounded Transformer (VGGT) has advanced 3D vision, yet its global attention layers suffer from quadratic computational costs that hinder scalability. Several spars…
Balancing Saliency and Coverage: Semantic Prominence-Aware Budgeting for Visual Token Compression in VLMs
Jaehoon Lee, Mingi Jung, Soohyuk Jang +3
Large Vision-Language Models (VLMs) achieve strong multimodal understanding capabilities by leveraging high-resolution visual inputs, but the resulting large number of visual token…
SAVE: Sparse Autoencoder-Driven Visual Information Enhancement for Mitigating Object Hallucination
Sangha Park, Seungryong Yoo, Jisoo Mok +1
Although Multimodal Large Language Models (MLLMs) have advanced substantially, they remain vulnerable to object hallucination caused by language priors and visual information loss.…
Guiding What Not to Generate: Automated Negative Prompting for Text-Image Alignment
Sangha Park, Eunji Kim, Yeongtak Oh +2
Despite substantial progress in text-to-image generation, achieving precise text-image alignment remains challenging, particularly for prompts with rich compositional structure or…
DefectFill: Realistic Defect Generation with Inpainting Diffusion Model for Visual Inspection
Jaewoo Song, Daemin Park, Kanghyun Baek +4
Developing effective visual inspection models remains challenging due to the scarcity of defect data. While image generation models have been used to synthesize defect images, prod…
Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models
Mingi Jung, Saehyung Lee, Eunji Kim +1
Detailed image captioning is essential for tasks like data generation and aiding visually impaired individuals. High-quality captions require a balance between precision and recall…
Superpixel Tokenization for Vision Transformers: Preserving Semantic Integrity in Visual Tokens
Jaihyun Lew, Soohyuk Jang, Jaehoon Lee +6
Transformers, a groundbreaking architecture proposed for Natural Language Processing (NLP), have also achieved remarkable success in Computer Vision. A cornerstone of their success…
Interpretable Next-token Prediction via the Generalized Induction Head
Eunji Kim, Sriya Mantena, Weiwei Yang +3
While large transformer models excel in predictive performance, their lack of interpretability restricts their usefulness in high-stakes domains. To remedy this, we propose the Gen…
Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP
Eunji Kim, Kyuhong Shim, Simyung Chang +1
A text encoder within Vision-Language Models (VLMs) like CLIP plays a crucial role in translating textual input into an embedding space shared with images, thereby facilitating the…