Temporally-Weighted Hierarchical Clustering for Unsupervised Action Segmentation

Action segmentation refers to inferring boundaries of semantically consistent visual concepts in videos and is an important requirement for many video understanding tasks. For this and other video understanding tasks, supervised approaches have achieved encouraging performance but require a high volume of detailed frame-level annotations. We present a fully automatic and unsupervised approach for segmenting actions in a video that does not require any training. Our proposal is an effective temporally-weighted hierarchical clustering algorithm that can group semantically consistent frames of the video. Our main finding is that representing a video with a 1-nearest neighbor graph by taking into account the time progression is sufficient to form semantically and temporally consistent clusters of frames where each cluster may represent some action in the video. Additionally, we establish strong unsupervised baselines for action segmentation and show significant performance improvements over published unsupervised methods on five challenging action segmentation datasets. Our code is available at https://github.com/ssarfraz/FINCH-Clustering/tree/master/TW-FINCH

PDF Abstract CVPR 2021 PDF CVPR 2021 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Action Segmentation 50 Salads TW-FINCH (K=avg/activity) Acc 66.5 # 27
Action Segmentation Breakfast TW-FINCH (K=avg/activity) Acc 62.7 # 33
mIoU 42.3 # 4
Action Segmentation MPII Cooking 2 Dataset Unsup. TW-FINCH (K=avg/activity) Accuracy 42 # 1
mIoU 23.1 # 1