computer-vision-7

Transductive Zero-Shot Classification

3260 papers • 126 benchmarks • 313 datasets

Zero-shot classification in the transductive setup where all test images are jointly available.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in transductive-zero-shot-classification-7

Trend

Dataset

Best Model

Actions

EuroSAT

ImageNet

DTD

Libraries

i

Use these libraries to find transductive-zero-shot-classification-7 models and implementations

Datasets

Subtasks

No subtasks available.

Most implemented papers

SSP: Self-Supervised Prompting for Cross-Lingual Transfer to Low-Resource Languages using Large Language Models

Parag Singla, Vipul Rathore, Aniruddha Deb, Ankish Chandresh, Mausam•Wed Jun 26 2024

This work introduces Self-Supervised Prompting (SSP), a novel ICL approach tailored for the 0-CLT setting that strongly outperforms existing SOTA fine-tuned and prompting-based baselines in 0-CLT setup and uses a novel Integer Linear Programming (ILP)-based exemplar selection that balances similarity, prediction confidence and label coverage.

Content

FGVC-Aircraft

Flowers-102

Food-101

Oxford-IIIT Pets

SUN397

Stanford Cars

Caltech-101

CIFAR-10

CIFAR-100

CUB-200-2011

UCF101

AID

RSC11

MLRSNet

OPTIMAL31

PatternNet

RESISC45

RSICB128

RSICB256

WHURS19

Oxford 102 Flower

2

0

Paper Graph

Label Propagation for Zero-shot Classification with Vision-Language Models

Giorgos Tolias, Yannis Kalantidis, Vladan Stojni'c•Thu Apr 04 2024

The case of zero-shot classification in the presence of unlabeled data is tackled and ZLaP, a method based on label propagation (LP) that utilizes geodesic distances for classification is introduced, a method based on label propagation that utilizes geodesic distances for classification.

27 0

Paper Graph

Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification

C. Vleeschouwer, Benoit G'erin, Ismail Ben Ayed, K. E. Khoury, Maxime Zanella, Tiffanie Godelaine, Benoit M. Macq, Saïd Mahmoudi•Sat Aug 31 2024

This work utilizes initial predictions based on text prompting and patch affinity relationships from the image encoder to enhance zero-shot capabilities through transductive inference, all without the need for supervision and at a minor computational cost.

8 0

Paper Graph

Adding a benchmark result helps the community track progress.