Inverse Neural Rendering for Explainable Multi-Object Tracking

arXiv:2404.12359v1 Announce Type: new
Abstract: Today, most methods for image understanding tasks rely on feed-forward neural networks. While this approach has allowed for empirical accuracy, efficiency, and task adaptation via fine-tuning, it also comes with fundamental disadvantages. Existing networks often struggle to generalize across different datasets, even on the same task. By design, these networks ultimately reason about high-dimensional scene features, which are challenging to analyze. This is true especially when attempting to predict 3D information based on 2D images. …

arxiv cs.cv cs.gr cs.ro neural rendering object rendering tracking type