论文标题
基于事件的数据的可区分复发表面
A Differentiable Recurrent Surface for Asynchronous Event-Based Data
论文作者
论文摘要
动态视觉传感器（DVSS）异步流中像素对应关系的流动事件会受到亮度的变化。与经典视觉设备不同，它们产生了场景的稀疏表示。因此，要应用标准的计算机视觉算法，需要将事件集成到框架或事件表面中。这通常是通过手工制作的网格来实现的，该网格使用临时启发式方法重建框架。在本文中，我们提出了Matrix-LSTM，即长期记忆（LSTM）单元格的网格，该网格有效地处理事件并学习端到端任务依赖性事件 - 曲面。与现有的重建方法相比，我们学到的事件表面对MVSEC基准的光流估算显示出良好的灵活性和表现力，并且它改善了N-CARS数据集中基于事件的对象分类的最先进。
Dynamic Vision Sensors (DVSs) asynchronously stream events in correspondence of pixels subject to brightness changes. Differently from classic vision devices, they produce a sparse representation of the scene. Therefore, to apply standard computer vision algorithms, events need to be integrated into a frame or event-surface. This is usually attained through hand-crafted grids that reconstruct the frame using ad-hoc heuristics. In this paper, we propose Matrix-LSTM, a grid of Long Short-Term Memory (LSTM) cells that efficiently process events and learn end-to-end task-dependent event-surfaces. Compared to existing reconstruction approaches, our learned event-surface shows good flexibility and expressiveness on optical flow estimation on the MVSEC benchmark and it improves the state-of-the-art of event-based object classification on the N-Cars dataset.
