Mimetics: Towards Understanding Human Actions Out of Context
This addresses the issue of context bias in action recognition for computer vision researchers, offering a new benchmark to evaluate true action understanding, though it is incremental as it builds on existing datasets and methods.
The paper tackles the problem of video action recognition models relying too heavily on contextual cues like scenes or objects, rather than understanding human actions themselves, by introducing the Mimetics dataset of mimed actions and showing that state-of-the-art 3D CNNs perform poorly on it, while models using body pose features achieve better results.
Recent methods for video action recognition have reached outstanding performances on existing benchmarks. However, they tend to leverage context such as scenes or objects instead of focusing on understanding the human action itself. For instance, a tennis field leads to the prediction playing tennis irrespectively of the actions performed in the video. In contrast, humans have a more complete understanding of actions and can recognize them without context. The best example of out-of-context actions are mimes, that people can typically recognize despite missing relevant objects and scenes. In this paper, we propose to benchmark action recognition methods in such absence of context and introduce a novel dataset, Mimetics, consisting of mimed actions for a subset of 50 classes from the Kinetics benchmark. Our experiments show that (a) state-of-the-art 3D convolutional neural networks obtain disappointing results on such videos, highlighting the lack of true understanding of the human actions and (b) models leveraging body language via human pose are less prone to context biases. In particular, we show that applying a shallow neural network with a single temporal convolution over body pose features transferred to the action recognition problem performs surprisingly well compared to 3D action recognition methods.