TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Zero-Shot Transfer Image Classification	aYahoo	CLIP	Accuracy	98.4	# 1
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Image-to-text R@1	58.4	# 10
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Image-to-text R@5	81.5	# 11
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Image-to-text R@10	88.1	# 10
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Text-to-image R@1	37.8	# 11
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Text-to-image R@5	62.4	# 11
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Text-to-image R@10	72.2	# 10
Zero-Shot Learning	COCO-MLT	ViT-B/16	Average mAP	60.17	# 2
Long-tail Learning	COCO-MLT	CLIP(ViT-B/16)	Average mAP	60.17	# 2
Zero-Shot Learning	COCO-MLT	ResNet-50	Average mAP	56.19	# 1
Long-tail Learning	COCO-MLT	CLIP(ResNet-50)	Average mAP	56.19	# 5
Prompt Engineering	EuroSAT	CLIP	Harmonic mean	60.03	# 2
Prompt Engineering	FGVC-Aircraft	CLIP	Harmonic mean	31.09	# 2
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Image-to-text R@1	88.0	# 9
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Image-to-text R@5	98.7	# 9
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Image-to-text R@10	99.4	# 8
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Text-to-image R@1	68.7	# 12
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Text-to-image R@5	90.6	# 12
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Text-to-image R@10	95.2	# 9
Object Categorization	GRIT	CLIP	Categorization (ablation)	48.1	# 3
Meme Classification	Hateful Memes	CLIP (zero-shot)	ROC-AUC	0.661	# 6
Prompt Engineering	ImageNet	CLIP	Harmonic mean	70.22	# 3
Zero-Shot Transfer Image Classification	ImageNet	CLIP (ResNet50)	Accuracy (Private)	59.6	# 16
Zero-Shot Transfer Image Classification	ImageNet	CLIP（ViT-L/14-336px）	Accuracy (Private)	76.2	# 12
Zero-Shot Transfer Image Classification	ImageNet	CLIP	Accuracy (Public)	31.3	# 2
Semi-Supervised Image Classification	ImageNet - 0.2% labeled data	CLIP (ResNet-50)	ImageNet Top-1 Accuracy	40%	# 3
Few-Shot Image Classification	ImageNet - 0-Shot	CLIP (ResNet50)	Accuracy	59.6%	# 3
Few-Shot Image Classification	ImageNet - 0-Shot	CLIP (ViT B/32)	Accuracy	63.2%	# 2
Prompt Engineering	ImageNet-A	CLIP	Top-1 accuracy %	47.77	# 5
Zero-Shot Transfer Image Classification	ImageNet-A	CLIP	Accuracy (Private)	77.2	# 7
Zero-Shot Transfer Image Classification	ImageNet-A	CLIP	Accuracy (Public)	-	# 2
Zero-Shot Transfer Image Classification	ImageNet-R	CLIP	Accuracy (Private)	88.9	# 8
Zero-Shot Transfer Image Classification	ImageNet-R	CLIP	Accuracy (Public)	-	# 2
Prompt Engineering	ImageNet-R	CLIP	Top-1 accuracy %	73.96	# 5
Prompt Engineering	ImageNet-S	CLIP	Top-1 accuracy %	46.15	# 5
Prompt Engineering	ImageNet V2	CLIP	Top-1 accuracy %	60.83	# 5
Zero-Shot Transfer Image Classification	ImageNet V2	CLIP	Accuracy (Private)	70.1	# 7
Zero-Shot Transfer Image Classification	ImageNet V2	CLIP	Accuracy (Public)	-	# 2
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-G/14, zero-shot, LAION-2B)	IN-W Gap	-3.8	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-G/14, zero-shot, LAION-2B)	Carton Gap	+12	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, WIT)	IN-W Gap	-4.4	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, WIT)	Carton Gap	+12	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, LAION-400M)	IN-W Gap	-4.9	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, LAION-400M)	Carton Gap	+12	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-H/14, zero-shot, LAION-2B)	IN-W Gap	-3.6	# 1
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-H/14, zero-shot, LAION-2B)	Carton Gap	+16	# 1
Zero-Shot Transfer Image Classification	ObjectNet	CLIP	Accuracy (Private)	72.3	# 5
Zero-Shot Transfer Image Classification	ObjectNet	CLIP	Accuracy (Public)	-	# 2
Image Classification	ObjectNet	CLIP	Top-1 Accuracy	72.3	# 8
Image Classification	OmniBenchmark	CLIP-RN50	Average Top-1 Accuracy	42.1	# 6
Open Vocabulary Attribute Detection	OVAD-Box benchmark	CLIP VIT-B16	mean average precision	16.6	# 7
Action Recognition	RareAct	CLIP	mWAP	40.7	# 2
Zero-Shot Transfer Image Classification	SUN	CLIP	Accuracy	58.5	# 1
Prompt Engineering	SUN397	CLIP	Harmonic mean	72.23	# 4
Prompt Engineering	UCF101	CLIP	Harmonic mean	73.85	# 4
Long-tail Learning	VOC-MLT	CLIP(ResNet-50)	Average mAP	84.30	# 4
Zero-Shot Learning	VOC-MLT	CLIP(ResNet-50)	Average mAP	84.30	# 1
Zero-Shot Learning	VOC-MLT	CLIP(ViT-B/16)	Average mAP	85.77	# 2
Long-tail Learning	VOC-MLT	CLIP(ViT-B/16)	Average mAP	85.77	# 2

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-learning-on-coco-mlt)](https://paperswithcode.com/sota/zero-shot-learning-on-coco-mlt?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/out-of-distribution-generalization-on-1)](https://paperswithcode.com/sota/out-of-distribution-generalization-on-1?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-2)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-2?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-learning-on-voc-mlt)](https://paperswithcode.com/sota/zero-shot-learning-on-voc-mlt?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/long-tail-learning-on-coco-mlt)](https://paperswithcode.com/sota/long-tail-learning-on-coco-mlt?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-eurosat)](https://paperswithcode.com/sota/prompt-engineering-on-eurosat?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-fgvc-aircraft)](https://paperswithcode.com/sota/prompt-engineering-on-fgvc-aircraft?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-1)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-1?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/few-shot-image-classification-on-imagenet-0)](https://paperswithcode.com/sota/few-shot-image-classification-on-imagenet-0?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/action-recognition-on-rareact)](https://paperswithcode.com/sota/action-recognition-on-rareact?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/long-tail-learning-on-voc-mlt)](https://paperswithcode.com/sota/long-tail-learning-on-voc-mlt?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/object-categorization-on-grit)](https://paperswithcode.com/sota/object-categorization-on-grit?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-imagenet)](https://paperswithcode.com/sota/prompt-engineering-on-imagenet?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/semi-supervised-image-classification-on-16)](https://paperswithcode.com/sota/semi-supervised-image-classification-on-16?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-sun397)](https://paperswithcode.com/sota/prompt-engineering-on-sun397?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-ucf101)](https://paperswithcode.com/sota/prompt-engineering-on-ucf101?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-imagenet-a)](https://paperswithcode.com/sota/prompt-engineering-on-imagenet-a?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-imagenet-r)](https://paperswithcode.com/sota/prompt-engineering-on-imagenet-r?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-imagenet-s)](https://paperswithcode.com/sota/prompt-engineering-on-imagenet-s?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/prompt-engineering-on-imagenet-v2)](https://paperswithcode.com/sota/prompt-engineering-on-imagenet-v2?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-6)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-6?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/meme-classification-on-hateful-memes)](https://paperswithcode.com/sota/meme-classification-on-hateful-memes?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/image-classification-on-omnibenchmark)](https://paperswithcode.com/sota/image-classification-on-omnibenchmark?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-5)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-5?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-3)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-3?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/open-vocabulary-attribute-detection-on-ovad-1)](https://paperswithcode.com/sota/open-vocabulary-attribute-detection-on-ovad-1?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-transfer-image-classification-on-4)](https://paperswithcode.com/sota/zero-shot-transfer-image-classification-on-4?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/image-classification-on-objectnet)](https://paperswithcode.com/sota/image-classification-on-objectnet?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-cross-modal-retrieval-on-flickr30k)](https://paperswithcode.com/sota/zero-shot-cross-modal-retrieval-on-flickr30k?p=learning-transferable-visual-models-from)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-transferable-visual-models-from/zero-shot-cross-modal-retrieval-on-coco-2014)](https://paperswithcode.com/sota/zero-shot-cross-modal-retrieval-on-coco-2014?p=learning-transferable-visual-models-from)`

Learning Transferable Visual Models From Natural Language Supervision

26 Feb 2021 · Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever ·

State-of-the-art computer vision systems are trained to predict a fixed set of predetermined object categories. This restricted form of supervision limits their generality and usability since additional labeled data is needed to specify any other visual concept. Learning directly from raw text about images is a promising alternative which leverages a much broader source of supervision. We demonstrate that the simple pre-training task of predicting which caption goes with which image is an efficient and scalable way to learn SOTA image representations from scratch on a dataset of 400 million (image, text) pairs collected from the internet. After pre-training, natural language is used to reference learned visual concepts (or describe new ones) enabling zero-shot transfer of the model to downstream tasks. We study the performance of this approach by benchmarking on over 30 different existing computer vision datasets, spanning tasks such as OCR, action recognition in videos, geo-localization, and many types of fine-grained object classification. The model transfers non-trivially to most tasks and is often competitive with a fully supervised baseline without the need for any dataset specific training. For instance, we match the accuracy of the original ResNet-50 on ImageNet zero-shot without needing to use any of the 1.28 million training examples it was trained on. We release our code and pre-trained model weights at https://github.com/OpenAI/CLIP.

PDF Abstract

Code

Add Remove Mark official

openai/CLIP official

↳ Quickstart in

Colab

17,839

mlfoundations/open_clip

↳ Quickstart in

Colab

6,242

facebookresearch/vissl

↳ Quickstart in

Colab

3,123

towhee-io/towhee

2,686

alibaba/EasyNLP

1,779

See all 40 implementations

Tasks

Add Remove

Action Recognition

Benchmarking

Few-Shot Image Classification

Image Classification

Long-tail Learning

Meme Classification

Natural Language Understanding

Object Categorization

Open Vocabulary Attribute Detection

Out-of-Distribution Generalization

Prompt Engineering

Semi-Supervised Image Classification

Text Generation

Zero-Shot Cross-Modal Retrieval

Zero-Shot Learning

Zero-Shot Transfer Image Classification

Datasets

ImageNet

COCO

KITTI

GLUE

SST

UCF101

Visual Genome

Flickr30k

FGVC-Aircraft

EuroSAT

WebText

ImageNet-A

ImageNet-R

YFCC100M

ImageNet-Sketch

FER2013 FairFace

Hateful Memes

ObjectNet

RESISC45

Kinetics-700

PCam

SUN397

OmniBenchmark

ImageNet-S

ImageNet-W

OVAD benchmark COCO-MLT VOC-MLT

GRIT

RareAct

Results from the Paper

Edit

Ranked #1 on Zero-Shot Learning on COCO-MLT (using extra training data)

Get a GitHub badge

Task	Dataset	Model	Metric Name	Metric Value	Global Rank	Benchmark
Zero-Shot Transfer Image Classification	aYahoo	CLIP	Accuracy	98.4	# 1	Compare
Zero-Shot Cross-Modal Retrieval	COCO 2014	CLIP	Image-to-text R@1	58.4	# 10	Compare
			Image-to-text R@5	81.5	# 11	Compare
			Image-to-text R@10	88.1	# 10	Compare
			Text-to-image R@1	37.8	# 11	Compare
			Text-to-image R@5	62.4	# 11	Compare
			Text-to-image R@10	72.2	# 10	Compare
Zero-Shot Learning	COCO-MLT	ViT-B/16	Average mAP	60.17	# 2	Compare
Long-tail Learning	COCO-MLT	CLIP(ViT-B/16)	Average mAP	60.17	# 2	Compare
Zero-Shot Learning	COCO-MLT	ResNet-50	Average mAP	56.19	# 1	Compare
Long-tail Learning	COCO-MLT	CLIP(ResNet-50)	Average mAP	56.19	# 5	Compare
Prompt Engineering	EuroSAT	CLIP	Harmonic mean	60.03	# 2	Compare
Prompt Engineering	FGVC-Aircraft	CLIP	Harmonic mean	31.09	# 2	Compare
Zero-Shot Cross-Modal Retrieval	Flickr30k	CLIP	Image-to-text R@1	88.0	# 9	Compare
			Image-to-text R@5	98.7	# 9	Compare
			Image-to-text R@10	99.4	# 8	Compare
			Text-to-image R@1	68.7	# 12	Compare
			Text-to-image R@5	90.6	# 12	Compare
			Text-to-image R@10	95.2	# 9	Compare
Object Categorization	GRIT	CLIP	Categorization (ablation)	48.1	# 3	Compare
Meme Classification	Hateful Memes	CLIP (zero-shot)	ROC-AUC	0.661	# 6	Compare
Prompt Engineering	ImageNet	CLIP	Harmonic mean	70.22	# 3	Compare
Zero-Shot Transfer Image Classification	ImageNet	CLIP (ResNet50)	Accuracy (Private)	59.6	# 16	Compare
Zero-Shot Transfer Image Classification	ImageNet	CLIP（ViT-L/14-336px）	Accuracy (Private)	76.2	# 12	Compare
Zero-Shot Transfer Image Classification	ImageNet	CLIP	Accuracy (Public)	31.3	# 2	Compare
Semi-Supervised Image Classification	ImageNet - 0.2% labeled data	CLIP (ResNet-50)	ImageNet Top-1 Accuracy	40%	# 3	Compare
Few-Shot Image Classification	ImageNet - 0-Shot	CLIP (ResNet50)	Accuracy	59.6%	# 3	Compare
Few-Shot Image Classification	ImageNet - 0-Shot	CLIP (ViT B/32)	Accuracy	63.2%	# 2	Compare
Prompt Engineering	ImageNet-A	CLIP	Top-1 accuracy %	47.77	# 5	Compare
Zero-Shot Transfer Image Classification	ImageNet-A	CLIP	Accuracy (Private)	77.2	# 7	Compare
Zero-Shot Transfer Image Classification	ImageNet-A	CLIP	Accuracy (Public)	-	# 2	Compare
Zero-Shot Transfer Image Classification	ImageNet-R	CLIP	Accuracy (Private)	88.9	# 8	Compare
Zero-Shot Transfer Image Classification	ImageNet-R	CLIP	Accuracy (Public)	-	# 2	Compare
Prompt Engineering	ImageNet-R	CLIP	Top-1 accuracy %	73.96	# 5	Compare
Prompt Engineering	ImageNet-S	CLIP	Top-1 accuracy %	46.15	# 5	Compare
Prompt Engineering	ImageNet V2	CLIP	Top-1 accuracy %	60.83	# 5	Compare
Zero-Shot Transfer Image Classification	ImageNet V2	CLIP	Accuracy (Private)	70.1	# 7	Compare
Zero-Shot Transfer Image Classification	ImageNet V2	CLIP	Accuracy (Public)	-	# 2	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-G/14, zero-shot, LAION-2B)	IN-W Gap	-3.8	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-G/14, zero-shot, LAION-2B)	Carton Gap	+12	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, WIT)	IN-W Gap	-4.4	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, WIT)	Carton Gap	+12	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, LAION-400M)	IN-W Gap	-4.9	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-L/14, zero-shot, LAION-400M)	Carton Gap	+12	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-H/14, zero-shot, LAION-2B)	IN-W Gap	-3.6	# 1	Compare
Out-of-Distribution Generalization	ImageNet-W	CLIP (ViT-H/14, zero-shot, LAION-2B)	Carton Gap	+16	# 1	Compare
Zero-Shot Transfer Image Classification	ObjectNet	CLIP	Accuracy (Private)	72.3	# 5	Compare
Zero-Shot Transfer Image Classification	ObjectNet	CLIP	Accuracy (Public)	-	# 2	Compare
Image Classification	ObjectNet	CLIP	Top-1 Accuracy	72.3	# 8	Compare
Image Classification	OmniBenchmark	CLIP-RN50	Average Top-1 Accuracy	42.1	# 6	Compare
Open Vocabulary Attribute Detection	OVAD-Box benchmark	CLIP VIT-B16	mean average precision	16.6	# 7	Compare
Action Recognition	RareAct	CLIP	mWAP

Edit Social Preview

Learning Transferable Visual Models From Natural Language Supervision

Code Edit Add Remove Mark official

Tasks Edit Add Remove

Datasets Edit

Results from the Paper Edit

Code

Add Remove Mark official

Tasks

Add Remove

Datasets

Results from the Paper

Edit