HIGHLIGHTS
- who: Yanan Liu and Walterio Mayol-Cuevas from the Peking University, China have published the paper: On-sensor binarized CNN inference with dynamic model swapping in pixel processor arrays, in the Journal: (JOURNAL)
- what: In the experiments, the authors implement the proposed CNN tree architecture on the sensor and demonstrate it with: real-time 37class English letter classification, objects` localization, and coarse image segmentation. The authors propose a dynamic-swapping CNN architecture where multiple neural_networks can be composed for more sophisticated inference tasks by dynamically uploading neural_network models when needed. With these challenges in mind . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.