Awesome

cEOT

Continuous-time overlap tracker for event data from Neuromorphic Vision Sensors.

The system takes as input asynchronous events defined by x and y coordinates, and a timestamp, and outputs bounding boxes defined by a center coordinate and its width and height.

Usage

First, add to path the MATLAB AER Vision Functions from Garrick Orchard.

Then, run the cEOT function:

track_cEOT('test_file', 0, 0);

The output of the process is an annotation file in text format.

An example script is included along with a short sample recording in bin format.

Citation

Andres Ussa, Chockalingam Senthil Rajen, Deepak Singla, Jyotibdha Acharya, Gideon Fu Chuanrong, Arindam Basu and Bharath Ramesh, "A Hybrid Neuromorphic Object Tracking and Classification Framework for Real-time Systems," arXiv:2007.11404 [cs.CV], 2020.

@misc{ussa2020hybrid,
    title={A Hybrid Neuromorphic Object Tracking and Classification Framework for Real-time Systems},
    author={Andres Ussa and Chockalingam Senthil Rajen and Deepak Singla and Jyotibdha Acharya and Gideon Fu Chuanrong and Arindam Basu and Bharath Ramesh},
    year={2020},
    eprint={2007.11404},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}