Table 1 Comparison of ANNs with ONNs for various problems
From: Computing with oscillators from theoretical underpinnings to applications and demonstrators
Problem | ANNs | Processing unit | ONNs |
|---|---|---|---|
Pattern recognition | Developed by Wu and He194 | GPU (GeForce GTX1080) | Developed by Zhang et al.118, Biswas et al.119, and Yun et al.120 |
Classification | Developed by Byerly et al.195 (99.83% on MNIST) | GPU (GeForce GTX1080 Tis) | Developed by Abernot and Todri-Sanial121 (~62% on MNIST), ClassONN122 (~70% on MNIST), 8 × 8 MNIST with EP179 (94.1%) |
Edge detection | DexiNed196 | GPU (Titan XP) | |
Online learning | Developed by Nose et al.197 | Raspberry Pi 3 Model B+ | |
Sensory processing | Developed by Jain et al.198 | GPU (K40) | Developed by Yang et al.127 |