Unsupervised Video Understanding by Reconciliation of Posture Similarities
This addresses the challenge of scalable and detailed human activity analysis for applications like video understanding, but it is incremental as it builds on unsupervised deep learning methods.
The paper tackles the problem of unsupervised fine-grained human activity understanding by learning a structured representation of postures and their transitions from video sequences without manual annotation, achieving capabilities such as posture retrieval, temporal super-resolution, and next-frame synthesis.
Understanding human activity and being able to explain it in detail surpasses mere action classification by far in both complexity and value. The challenge is thus to describe an activity on the basis of its most fundamental constituents, the individual postures and their distinctive transitions. Supervised learning of such a fine-grained representation based on elementary poses is very tedious and does not scale. Therefore, we propose a completely unsupervised deep learning procedure based solely on video sequences, which starts from scratch without requiring pre-trained networks, predefined body models, or keypoints. A combinatorial sequence matching algorithm proposes relations between frames from subsets of the training data, while a CNN is reconciling the transitivity conflicts of the different subsets to learn a single concerted pose embedding despite changes in appearance across sequences. Without any manual annotation, the model learns a structured representation of postures and their temporal development. The model not only enables retrieval of similar postures but also temporal super-resolution. Additionally, based on a recurrent formulation, next frames can be synthesized.