Accuracy (Cross-Subject): 88.7Accuracy (Cross-Setup): 89.1
Papers with Code · CC BY-SA 4.0
Zero-Shot Learning
Zero-shot learning (ZSL) is a model's ability to detect classes never seen during training. The condition is that the classes are not known during supervised learning.
Earlier work in zero-shot learning use attributes in a two-step approach to infer unknown classes. In the computer vision context, more recent advances learn mappings from image feature space to semantic space. Other approaches learn non-linear multimodal embeddings. In the modern NLP context, language models can be evaluated on downstream tasks without fine tuning.
Benchmark datasets for zero-shot learning include aPY, AwA, and CUB, among others.
( Image credit: Prototypical Networks for Few shot Learning in PyTorch )
Further readings:
논문
Accuracy (Cross-Subject): 89.7Accuracy (Cross-Setup): 91.0
Accuracy (Cross-Subject): 88.9Accuracy (Cross-Setup): 90.6Ensembled Modalities: 4
Accuracy (Cross-Subject): 89.6Accuracy (Cross-Setup): 91.3Ensembled Modalities: 4
Accuracy (Cross-Subject): 86.9%Accuracy (Cross-Setup): 88.4%
Accuracy (Cross-Subject): 89.12Accuracy (Cross-Setup): 90.97Ensembled Modalities: 4
Accuracy (Cross-Subject): 88.2%Accuracy (Cross-Setup): 89.2%Ensembled Modalities: 4
Accuracy (Cross-Subject): 89.9Accuracy (Cross-Setup): 91.1Ensembled Modalities: 4
Accuracy (Cross-Subject): 89.5Accuracy (Cross-Setup): 91.0Ensembled Modalities: 4
Accuracy (Cross-Subject): 90.1Accuracy (Cross-Setup): 91.6Ensembled Modalities: 6
Accuracy (Cross-Subject): 89.9Accuracy (Cross-Setup): 91.3Ensembled Modalities: 4
Accuracy (Cross-Subject): 89.8Accuracy (Cross-Setup): 91.2Ensembled Modalities: 6
Accuracy (Cross-Subject): 89.4Accuracy (Cross-Setup): 91.4Ensembled Modalities: 4
Accuracy (Cross-Subject): 88.2Accuracy (Cross-Setup): 89.3Ensembled Modalities: 4
Accuracy (Cross-Subject): 87.5Accuracy (Cross-Setup): 89.2Ensembled Modalities: 4
Accuracy (Cross-Subject): 88.7Accuracy (Cross-Setup): 91.5
Accuracy (Cross-Subject): 87.4Accuracy (Cross-Setup): 89.5
Accuracy (Cross-Subject): 90.0Accuracy (Cross-Setup): 91.3
Accuracy (Cross-Subject): 87.4Accuracy (Cross-Setup): 88.3
Accuracy (Cross-Subject): 89.4Accuracy (Cross-Setup): 90.6
Accuracy (Cross-Subject): 88.6Accuracy (Cross-Setup): 90.8Ensembled Modalities: 4
Accuracy (Cross-Subject): 87.22%Accuracy (Cross-Setup): 88.8%
Accuracy (Cross-Subject): 86.9Accuracy (Cross-Setup): 90.3
Accuracy (Cross-Subject): 89.8Accuracy (Cross-Setup): 91.4Ensembled Modalities: 4
Accuracy (Cross-Subject): 89.4Accuracy (Cross-Setup): 90.8Ensembled Modalities: 4
Accuracy (Cross-Subject): 88.7Accuracy (Cross-Setup): 90.4Ensembled Modalities: 4
Accuracy (Cross-Subject): 90.0Accuracy (Cross-Setup): 91.2Ensembled Modalities: 4
Accuracy (Cross-Subject): 87.3Accuracy (Cross-Setup): 88.3
Accuracy (Cross-Subject): 89.1Accuracy (Cross-Setup): 90.3Ensembled Modalities: 4