ActionFlowNet: Learning Motion Representation for Action Recognition
This addresses the need for improved motion representation in action recognition for computer vision applications, offering a novel integrated approach that reduces reliance on handcrafted features.
The paper tackles the problem of action recognition by proposing ActionFlowNet, a multitask learning model that jointly estimates optical flow and recognizes actions from raw pixels, achieving a 31% accuracy improvement over state-of-the-art CNN-based models without external data or optical flow input.
Even with the recent advances in convolutional neural networks (CNN) in various visual recognition tasks, the state-of-the-art action recognition system still relies on hand crafted motion feature such as optical flow to achieve the best performance. We propose a multitask learning model ActionFlowNet to train a single stream network directly from raw pixels to jointly estimate optical flow while recognizing actions with convolutional neural networks, capturing both appearance and motion in a single model. We additionally provide insights to how the quality of the learned optical flow affects the action recognition. Our model significantly improves action recognition accuracy by a large margin 31% compared to state-of-the-art CNN-based action recognition models trained without external large scale data and additional optical flow input. Without pretraining on large external labeled datasets, our model, by well exploiting the motion information, achieves competitive recognition accuracy to the models trained with large labeled datasets such as ImageNet and Sport-1M.