TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	86.99	# 8
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	87.16	# 7
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	76.2	# 21
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	75.24	# 24
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	90.47	# 10
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	90.2	# 11
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	88.38	# 19
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	89.0	# 15
Few-Shot Image Classification	CUB 200 5-way	EASY 3xResNet12 (transductive)	Accuracy	93.79	# 1
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 3xResNet12 (transductive)	Accuracy	90.56	# 8
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 4xResNet12 (transductive)	Accuracy	90.5	# 9
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 3xResNet12 (inductive)	Accuracy	78.56	# 20
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 4xResNet12 (inductive)	Accuracy	77.97	# 21
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 4xResNet12 (transductive)	Accuracy	93.5	# 7
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 3xResNet12 (inductive)	Accuracy	91.93	# 12
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 4xResNet12 (inductive)	Accuracy	91.59	# 13
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	47.94	# 9
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	54.47	# 2
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	54.13	# 3
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	48.07	# 8
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	64.74	# 9
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	66.86	# 5
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	64.14	# 10
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	65.82	# 7
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	71.75	# 27
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	70.63	# 30
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	82.31	# 11
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	84.04	# 8
Few-Shot Learning	Mini-Imagenet 5-way (1-shot)	EASY (transductive)	Accuracy	82.75	# 1
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	89.14	# 11
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	88.57	# 14
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	86.28	# 19
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	87.15	# 16
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	ASY ResNet12 (transductive)	Accuracy	83.98	# 8
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	84.29	# 6
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	74.71	# 19
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	ASY ResNet12 (ours)	Accuracy	74.31	# 21
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	88.33	# 15
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	ASY ResNet12 (inductive)	Accuracy	87.86	# 17
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	89.76	# 9
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	ASY ResNet12 (transductive)	Accuracy	89.26	# 11

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-cub-200-5-2)](https://paperswithcode.com/sota/few-shot-image-classification-on-cub-200-5-2?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-learning-on-mini-imagenet-5-way-1)](https://paperswithcode.com/sota/few-shot-learning-on-mini-imagenet-5-way-1?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-fc100-5-way)](https://paperswithcode.com/sota/few-shot-image-classification-on-fc100-5-way?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-fc100-5-way-1)](https://paperswithcode.com/sota/few-shot-image-classification-on-fc100-5-way-1?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-tiered)](https://paperswithcode.com/sota/few-shot-image-classification-on-tiered?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-cifar-fs-5)](https://paperswithcode.com/sota/few-shot-image-classification-on-cifar-fs-5?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-cub-200-5)](https://paperswithcode.com/sota/few-shot-image-classification-on-cub-200-5?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-cub-200-5-1)](https://paperswithcode.com/sota/few-shot-image-classification-on-cub-200-5-1?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-mini-2)](https://paperswithcode.com/sota/few-shot-image-classification-on-mini-2?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-tiered-1)](https://paperswithcode.com/sota/few-shot-image-classification-on-tiered-1?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-cifar-fs-5-1)](https://paperswithcode.com/sota/few-shot-image-classification-on-cifar-fs-5-1?p=easy-ensemble-augmented-shot-y-shaped)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/easy-ensemble-augmented-shot-y-shaped/few-shot-image-classification-on-mini-3)](https://paperswithcode.com/sota/few-shot-image-classification-on-mini-3?p=easy-ensemble-augmented-shot-y-shaped)`

EASY: Ensemble Augmented-Shot Y-shaped Learning: State-Of-The-Art Few-Shot Classification with Simple Ingredients

24 Jan 2022 · Yassir Bendou, Yuqing Hu, Raphael Lafargue, Giulia Lioi, Bastien Pasdeloup, Stéphane Pateux, Vincent Gripon ·

Few-shot learning aims at leveraging knowledge learned by one or more deep learning models, in order to obtain good classification performance on new problems, where only a few labeled samples per class are available. Recent years have seen a fair number of works in the field, introducing methods with numerous ingredients. A frequent problem, though, is the use of suboptimally trained models to extract knowledge, leading to interrogations on whether proposed approaches bring gains compared to using better initial models without the introduced ingredients. In this work, we propose a simple methodology, that reaches or even beats state of the art performance on multiple standardized benchmarks of the field, while adding almost no hyperparameters or parameters to those used for training the initial deep learning models on the generic dataset. This methodology offers a new baseline on which to propose (and fairly compare) new techniques or adapt existing ones.

PDF Abstract

Code

Add Remove Mark official

ybendou/easy official

104

ybendou/fs-generalization

Tasks

Add Remove

Few-Shot Image Classification

Few-Shot Learning

Datasets

CUB-200-2011 mini-Imagenet

tieredImageNet

CIFAR-FS

FC100 miniImageNet

Results from the Paper

Edit

Ranked #1 on Few-Shot Learning on Mini-Imagenet 5-way (1-shot)

Get a GitHub badge

Task	Dataset	Model	Metric Name	Metric Value	Global Rank	Benchmark
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	86.99	# 8	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	87.16	# 7	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	76.2	# 21	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	75.24	# 24	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	90.47	# 10	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	90.2	# 11	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	88.38	# 19	Compare
Few-Shot Image Classification	CIFAR-FS 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	89.0	# 15	Compare
Few-Shot Image Classification	CUB 200 5-way	EASY 3xResNet12 (transductive)	Accuracy	93.79	# 1	Compare
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 3xResNet12 (transductive)	Accuracy	90.56	# 8	Compare
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 4xResNet12 (transductive)	Accuracy	90.5	# 9	Compare
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 3xResNet12 (inductive)	Accuracy	78.56	# 20	Compare
Few-Shot Image Classification	CUB 200 5-way 1-shot	EASY 4xResNet12 (inductive)	Accuracy	77.97	# 21	Compare
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 4xResNet12 (transductive)	Accuracy	93.5	# 7	Compare
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 3xResNet12 (inductive)	Accuracy	91.93	# 12	Compare
Few-Shot Image Classification	CUB 200 5-way 5-shot	EASY 4xResNet12 (inductive)	Accuracy	91.59	# 13	Compare
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	47.94	# 9	Compare
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	54.47	# 2	Compare
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	54.13	# 3	Compare
Few-Shot Image Classification	FC100 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	48.07	# 8	Compare
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	64.74	# 9	Compare
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	66.86	# 5	Compare
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	64.14	# 10	Compare
Few-Shot Image Classification	FC100 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	65.82	# 7	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	71.75	# 27	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	70.63	# 30	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	82.31	# 11	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	84.04	# 8	Compare
Few-Shot Learning	Mini-Imagenet 5-way (1-shot)	EASY (transductive)	Accuracy	82.75	# 1	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	89.14	# 11	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 2xResNet12 1/√2 (transductive)	Accuracy	88.57	# 14	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 2xResNet12 1/√2 (inductive)	Accuracy	86.28	# 19	Compare
Few-Shot Image Classification	Mini-Imagenet 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	87.15	# 16	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	ASY ResNet12 (transductive)	Accuracy	83.98	# 8	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	EASY 3xResNet12 (transductive)	Accuracy	84.29	# 6	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	EASY 3xResNet12 (inductive)	Accuracy	74.71	# 19	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (1-shot)	ASY ResNet12 (ours)	Accuracy	74.31	# 21	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	EASY 3xResNet12 (inductive)	Accuracy	88.33	# 15	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	ASY ResNet12 (inductive)	Accuracy	87.86	# 17	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	EASY 3xResNet12 (transductive)	Accuracy	89.76	# 9	Compare
Few-Shot Image Classification	Tiered ImageNet 5-way (5-shot)	ASY ResNet12 (transductive)	Accuracy	89.26	# 11	Compare

Methods

Add Remove

No methods listed for this paper. Add relevant methods here

Edit Social Preview

EASY: Ensemble Augmented-Shot Y-shaped Learning: State-Of-The-Art Few-Shot Classification with Simple Ingredients

Code Edit Add Remove Mark official

Tasks Edit Add Remove

Datasets Edit

Results from the Paper Edit

Methods Edit Add Remove

Code

Add Remove Mark official

Tasks

Add Remove

Datasets

Results from the Paper

Edit

Methods

Add Remove