Distribution Auto-Encoder for Action Quality Assessment from Videos
Auto-Encoding Score Distribution Regression for Action Quality Assessment
Action quality assessment (aqa) from videos is a challenging vision task since the relation between videos and action scores is difficult to model.
Traditionally, aqa task is treated as a regression problem to learn the underlying mappings between videos and action scores.
More recently, the methodof uncertainty score distribution learning (usdl) made success due to theintroduction of label distribution learning (ldl).
But usdl does not apply todataset with continuous labels and needs a fixed variance in training.
In this paper, to address the above problems, we further develop distributionauto-encoder (dae).
Specifically, it encodes videos into distributions and uses the reparameterization trick in variational auto-encoders (vae) to sample scores, which establishes a more accurate mapping between videos and scores.
Meanwhile, a combined loss is constructed to accelerate the training of dae.
Experimental results on public datasets demonstrate that our method achieves state-of-the-arts under the spearman s rank correlation : 0.9449 on multi-task datasets and 0.73 on multi-task datasets.