#dense prediction

try —

4 papers match

cs.CV2026

U-shaped Multi-granularity Learning for Vision-Language Models

Biao Chen, Yunqian Yu, Xiangxu Zhao +3

The paper introduces UPrompt, a U‑shaped multi‑granularity prompt learning framework that combines global and local prompts for vision‑language models, improving fine‑grained seman…

#prompt learning#vision-language models#multi-granularity#dense prediction
cs.CV2026

Panoramic Scene Understanding: A Survey from Distortion-Aware Engineering to Sphere-Native Modeling

Qinfeng Zhu, Lei Fan

The paper surveys methods for understanding panoramic images, covering techniques that handle projection distortions, sphere-native models, and adaptations of pretrained vision fou…

#panoramic imaging#spherical geometry#distortion-aware modeling#dense prediction
cs.CV2026

Let RGB Be the Language of Vision

Timing Yang, Jinrui Yang, Xinlong Li +11

The paper proposes a unified vision framework that encodes all visual signals—including images, masks, and depth maps—as RGB images, turning diverse tasks into a common RGB-to-RGB…

#unified vision model#rgb representation#image editing#zero-shot learning
cs.CV2026

Sparse Attention for Dense Open-Vocabulary Prediction in CLIP

Fatimah Zohra, Chen Zhao, Shuming Liu +1

The paper replaces the softmax in CLIP's visual self‑attention with the α‑entmax transform to create sparse attention, which reduces noise from irrelevant tokens and improves dense…

#sparse attention#open-vocabulary segmentation#entmax#contrastive language-image pretraining