Learning to Predict Diverse Human Motions from a Single Image via Mixture Density Networks
This addresses the challenge of motion prediction in computer vision for real-world applications where obtaining complete past sequences is costly, though it is incremental as it builds on existing MDN and energy-based methods.
The paper tackles the problem of predicting diverse future human motions from a single image, a weaker condition than typical sequence-based inputs, using mixture density networks (MDN) and achieves improved diversity and accuracy on standard benchmarks.
Human motion prediction, which plays a key role in computer vision, generally requires a past motion sequence as input. However, in real applications, a complete and correct past motion sequence can be too expensive to achieve. In this paper, we propose a novel approach to predicting future human motions from a much weaker condition, i.e., a single image, with mixture density networks (MDN) modeling. Contrary to most existing deep human motion prediction approaches, the multimodal nature of MDN enables the generation of diverse future motion hypotheses, which well compensates for the strong stochastic ambiguity aggregated by the single input and human motion uncertainty. In designing the loss function, we further introduce the energy-based formulation to flexibly impose prior losses over the learnable parameters of MDN to maintain motion coherence as well as improve the prediction accuracy by customizing the energy functions. Our trained model directly takes an image as input and generates multiple plausible motions that satisfy the given condition. Extensive experiments on two standard benchmark datasets demonstrate the effectiveness of our method in terms of prediction diversity and accuracy.