Anchor-Constrained Viterbi for Set-Supervised Action Segmentation
This addresses the problem of reducing annotation costs for video action segmentation, but it is incremental as it builds on existing HMM and weak supervision methods.
The paper tackles action segmentation under weak supervision, where only action sets are provided without temporal ordering, by introducing an anchor-constrained Viterbi algorithm to generate pseudo-ground truth, achieving superior performance on benchmark datasets like Breakfast, MPII Cooking2, and Hollywood Extended.
This paper is about action segmentation under weak supervision in training, where the ground truth provides only a set of actions present, but neither their temporal ordering nor when they occur in a training video. We use a Hidden Markov Model (HMM) grounded on a multilayer perceptron (MLP) to label video frames, and thus generate a pseudo-ground truth for the subsequent pseudo-supervised training. In testing, a Monte Carlo sampling of action sets seen in training is used to generate candidate temporal sequences of actions, and select the maximum posterior sequence. Our key contribution is a new anchor-constrained Viterbi algorithm (ACV) for generating the pseudo-ground truth, where anchors are salient action parts estimated for each action from a given ground-truth set. Our evaluation on the tasks of action segmentation and alignment on the benchmark Breakfast, MPII Cooking2, Hollywood Extended datasets demonstrates our superior performance relative to that of prior work.