SOTAVerified

Visual Prompting

Visual Prompting is the task of streamlining computer vision processes by harnessing the power of prompts, inspired by the breakthroughs of text prompting in NLP. This innovative approach involves using a few visual prompts to swiftly convert an unlabeled dataset into a deployed model, significantly reducing development time for both individual projects and enterprise solutions.

Papers

Showing 3140 of 127 papers

TitleStatusHype
Tune-An-Ellipse: CLIP Has Potential to Find What You WantCode1
Understanding and Improving Visual Prompting: A Label-Mapping PerspectiveCode1
Finding Visual Task VectorsCode1
GeoSAM: Fine-tuning SAM with Multi-Modal Prompts for Mobility Infrastructure SegmentationCode1
Token Coordinated Prompt Attention is Needed for Visual PromptingCode1
EZ-CLIP: Efficient Zeroshot Video Action RecognitionCode1
By My Eyes: Grounding Multimodal Large Language Models with Sensor Data via Visual PromptingCode1
Fine-Grained Visual PromptingCode1
Exploring the Transferability of Visual Prompting for Multimodal Large Language ModelsCode1
Improved GUI Grounding via Iterative NarrowingCode1
Show:102550
← PrevPage 4 of 13Next →

No leaderboard results yet.