Accuracy (CV): 70.27Accuracy (CS): 62.93
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 (RGB+pose): 86.1
Accuracy (RGB+pose): 90.4Accuracy (pose): 67.9
Accuracy (CV): 77.7Accuracy (CS): 69.2
mean per-class accuracy: 34.61
Accuracy (CV): 83.2Accuracy (CS): 74.6
Accuracy (CV): 81.2Accuracy (CS): 73.4
Accuracy (RGB+pose): 76.1Accuracy (pose): 56.8
mean per-class accuracy: 43.59
Accuracy (CV): 81.25Accuracy (CS): 74.60
Accuracy (RGB+pose): 62.5
10%: 58.1
Accuracy (CV): 64.0Accuracy (CS): 59.1
Accuracy (RGB+pose): -Accuracy (pose): 86.1
Accuracy (CV): 52.8Accuracy (CS): 50.1
Accuracy (CV): 83.1Accuracy (CS): 74.3
Accuracy (CV): 82.31
Accuracy (CV): 65.2Accuracy (CS): 60.2
Accuracy (CV): 79.5Accuracy (CS): 71.3
Accuracy (RGB+pose): 71.1
Accuracy (RGB+pose): 84.1
10%: 60.6
Accuracy (RGB+pose): 83.9
Accuracy (CV): 41.4Accuracy (CS): 38.6
Skeleton Image Representation for 3D Action Recognition based on Tree Structure and Reference Joints
Accuracy (CV): 80.3Accuracy (CS): 73.3
Accuracy (RGB+pose): 64.3