Temporally Coherent Embeddings for Self-Supervised Video Representation Learning

Open paper
Benchmark Model Rank Results
self-supervised-action-recognition-on-ucf101TCE (ResNet-50)#343-fold Accuracy: 71.2Pre-Training Dataset: Kinetics400Frozen: false
self-supervised-action-recognition-on-ucf101TCE (ResNet-18, Split 1)#353-fold Accuracy: 68.8Pre-Training Dataset: Kinetics400Frozen: false
self-supervised-action-recognition-on-ucf101TCE (ResNet18, Split 1)#373-fold Accuracy: 68.2Pre-Training Dataset: UCF101Frozen: false
self-supervised-action-recognition-on-hmdb51TCE (ResNet-50)#35Top-1 Accuracy: 36.6Pre-Training Dataset: Kinetics400Frozen: false
self-supervised-action-recognition-on-hmdb51TCE (ResNet-18)#38Top-1 Accuracy: 34.2Pre-Training Dataset: Kinetics400Frozen: false

Rank counts only results with a code link.