imagecompression1995/clip

collect recent papers on prompt learning

CLIP

Collect recent papers on CLIP and prompt learning

core thinking

• Why CLIP-prior works well?
  • CLIP learns relationships between vision and language from 400 million text-image pairs.
• How to transfer?
  • Zero-shot transfer
  • Prompt learning

Seminal Work

Title	Year	Venue	Code	Notes
Learning Transferable Visual Models From Natural Language Supervision	2021	ICML	Link	Link

Improved Work

Title	Year	Venue	Code
Supervision Exists Everywhere: A Data Efficient Contrastive Language-Image Pre-training Paradigm	2022	ICLR	Link
BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation	2022	Arxiv	Link
SLIP: Self-supervision meets Language-Image Pre-training	2022	Arxiv	Link

Applications

Image Classification

Title	Year	Venue	Code
Learning to Prompt for Vision-Language Models	2021	Arxiv	Link
Neural Prompt Search	2022	Arxiv	None
Prompt Distribution Learning	2022	CVPR	None
Conditional Prompt Learning for Vision-Language Models	2022	CVPR	Link
CLIP-Adapter: Better Vision-Language Models with Feature Adapters	2022	Arxiv	Link
Tip-Adapter: Training-free CLIP-Adapter for Better Vision-Language Modeling	2022	Arxiv	Link
Unsupervised Prompt Learning for Vision-Language Models	2022	Arxiv	Link

Detection

Title	Year	Venue	Code	Notes
Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model	2022	CVPR	Link	Link
DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting	2022	CVPR	Link	Link

Image/Video translation/editing

Title	Year	Venue	Code
HairCLIP: Design Your Hair by Text and Reference Image	2022	CVPR	Link
FlexIT: Towards Flexible Semantic Image Translation	2022	CVPR	None
VQGAN-CLIP: Open Domain Image Generation and Editing with Natural Language Guidance	2022	Arxiv	Link

Image/Video Understanding

Title	Year	Venue	Code
Bridge-Prompt: Towards Ordinal Action Understanding in Instructional Videos	2022	CVPR	Link

Visual Grounding

Title	Year	Venue	Code
ClipCap: CLIP Prefix for Image Captioning	2021	Arxiv	Link
CPT: Colorful prompt tuning for pre-trained vision-language models	2021	Arxiv	None
ReCLIP: A Strong Zero-Shot Baseline for Referring Expression Comprehension	2022	ACL	None