Few-Shot Image Classification

Few-Shot Image Classification is a computer vision task that involves training machine learning models to classify images into predefined categories using only a few labeled examples of each category (typically ( Image credit: Learning Embedding Adaptation for Few-Shot Learning )

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 141–150 of 353 papers

Title	Date	Tasks	Status	Hype
Multi-level Second-order Few-shot Learning	Jan 15, 2022	Action RecognitionFew-Shot action recognition	CodeCode Available	0
HyperTransformer: Model Generation for Supervised and Semi-Supervised Few-Shot Learning	Jan 11, 2022	Few-Shot Image ClassificationFew-Shot Learning	CodeCode Available	1
Debiased Learning from Naturally Imbalanced Pseudo-Labels	Jan 5, 2022	counterfactualCounterfactual Reasoning	CodeCode Available	1
Improving Adversarially Robust Few-Shot Image Classification With Generalizable Representations	Jan 1, 2022	ClassificationFew-Shot Image Classification	—Unverified	0
Transformers Can Do Bayesian Inference	Dec 20, 2021	AutoMLBayesian Inference	CodeCode Available	1
Exploring Category-correlated Feature for Few-shot Image Classification	Dec 14, 2021	ClassificationDecoder	—Unverified	0
Few-Shot Image Classification Along Sparse Graphs	Dec 7, 2021	ClassificationFew-Shot Image Classification	—Unverified	0
Curriculum Meta-Learning for Few-shot Classification	Dec 6, 2021	ClassificationFew-Shot Image Classification	CodeCode Available	0
Inferring Prototypes for Multi-Label Few-Shot Image Classification with Word Vector Guided Attention	Dec 2, 2021	DescriptiveFew-Shot Image Classification	—Unverified	0
argmax centroid	Dec 1, 2021	Domain AdaptationFew-Shot Image Classification	—Unverified	0

Show:10 25 50

← PrevPage 15 of 36Next →

All datasets Mini-Imagenet 5-way (1-shot)Mini-Imagenet 5-way (5-shot)Tiered ImageNet 5-way (5-shot)Tiered ImageNet 5-way (1-shot)CIFAR-FS 5-way (5-shot)CIFAR-FS 5-way (1-shot)CUB 200 5-way 1-shot CUB 200 5-way 5-shot FC100 5-way (1-shot)FC100 5-way (5-shot)Meta-Dataset OMNIGLOT - 1-Shot, 20-way

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	SgVA-CLIP	Accuracy	97.95	—	Unverified
2	CAML [Laion-2b]	Accuracy	96.2	—	Unverified
3	P>M>F (P=DINO-ViT-base, M=ProtoNet)	Accuracy	95.3	—	Unverified
4	TRIDENT	Accuracy	86.11	—	Unverified
5	PT+MAP+SF+SOT (transductive)	Accuracy	85.59	—	Unverified
6	PT+MAP+SF+BPA (transductive)	Accuracy	85.59	—	Unverified
7	PEMnE-BMS* (transductive)	Accuracy	85.54	—	Unverified
8	PT+MAP (s+f) (transductive)	Accuracy	84.81	—	Unverified
9	BAVARDAGE	Accuracy	84.8	—	Unverified
10	EASY 3xResNet12 (transductive)	Accuracy	84.04	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	SgVA-CLIP	Accuracy	98.72	—	Unverified
2	CAML [Laion-2b]	Accuracy	98.6	—	Unverified
3	P>M>F (P=DINO-ViT-base, M=ProtoNet)	Accuracy	98.4	—	Unverified
4	TRIDENT	Accuracy	95.95	—	Unverified
5	BAVARDAGE	Accuracy	91.65	—	Unverified
6	PEMnE-BMS*(transductive)	Accuracy	91.53	—	Unverified
7	Transductive CNAPS + FETI	Accuracy	91.5	—	Unverified
8	PT+MAP+SF+BPA (transductive)	Accuracy	91.34	—	Unverified
9	PT+MAP+SF+SOT (transductive)	Accuracy	91.34	—	Unverified
10	AmdimNet	Accuracy	90.98	—	Unverified