Accuracy (CV): 94.8Accuracy (CS): 88.9GFLOPs per pred: 0.3
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 (CV): 95.2Accuracy (CS): 89.1
Accuracy (CV): 94.2Accuracy (CS): 86.2
Accuracy (CV): 94.2Accuracy (CS): 86.8
Accuracy (CV): 95.0Accuracy (CS): 89.2
Accuracy (CV): 95.5Accuracy (CS): 88.3
Accuracy (CV): 91.1Accuracy (CS): 86.5
Accuracy (CV): 93.97Accuracy (CS): 86.70
Learning Graph Convolutional Network for Skeleton-based Human Action Recognition by Neural Searching
Accuracy (CV): 95.7Accuracy (CS): 89.4
Accuracy (CV): 91.6Accuracy (CS): 86.8
Accuracy (CV): 95.1Accuracy (CS): 88.5
Accuracy (CV): 94.3Accuracy (CS): 87.5
Accuracy (CV): 93.2Accuracy (CS): 87.5
Accuracy (CV): 93.6Accuracy (CS): 88.0
Accuracy (CV): 93.5Accuracy (CS): 85.9
Accuracy (CV): 93.6Accuracy (CS): 87.3
Accuracy (CV): 92.7Accuracy (CS): 87.2
Accuracy (CV): 94.5Accuracy (CS): 89.0
Accuracy (CV): 96.0Accuracy (CS): 89.4
Accuracy (CV): 94.9Accuracy (CS): 89.1
Accuracy (CV): 93.2Accuracy (CS): 86.6
Accuracy (CV): 93.7Accuracy (CS): 88.6
USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation
Accuracy (CV): 93.2Accuracy (CS): 87.1
Accuracy (CV): 95.0Accuracy (CS): 89.4