Pose-Invariant Object Recognition for Event-Based Vision with Slow-ELM
This work addresses the lack of algorithms for capturing invariance to transformations in event-based vision, which is important for low-power, real-time systems like object recognition and pose-estimation.
The paper tackles the problem of pose-invariant object recognition in event-based vision by proposing a slow-ELM architecture that combines Extreme Learning Machines and Slow Feature Analysis, achieving 1% classification error for 8 objects over 90 degrees of 2D pose and performing 10,000 classifications per second on a CPU.
Neuromorphic image sensors produce activity-driven spiking output at every pixel. These low-power consuming imagers which encode visual change information in the form of spikes help reduce computational overhead and realize complex real-time systems; object recognition and pose-estimation to name a few. However, there exists a lack of algorithms in event-based vision aimed towards capturing invariance to transformations. In this work, we propose a methodology for recognizing objects invariant to their pose with the Dynamic Vision Sensor (DVS). A novel slow-ELM architecture is proposed which combines the effectiveness of Extreme Learning Machines and Slow Feature Analysis. The system, tested on an Intel Core i5-4590 CPU, can perform 10,000 classifications per second and achieves 1% classification error for 8 objects with views accumulated over 90 degrees of 2D pose.