Image Classification on ImageNet balanced original (test)
72.73AccuracyEmaQ-LT
Evaluation results
| Method | Links | |
|---|---|---|
| EmaQ-LTBackbone=ResNet-50, Bit-width=4 bits2026.06 | 72.73 | |
| SQuantBackbone=ResNet-50, Bit-width=4 bits2026.06 | 70.8 | |
| ZAQBackbone=ResNet-50, Bit-width=4 bits2026.06 | 70.06 | |
| ZeroQBackbone=ResNet-50, Bit-width=4 bits2026.06 | 69.3 | |
| Choi et al.Backbone=ResNet-50, Bit-width=4 bits2026.06 | 69.1 | |
| ARC+AITBackbone=ResNet-50, Bit-width=4 bits2026.06 | 68.27 | |
| BatchQuantBackbone=ResNet-50, Bit-width=4 bits2026.06 | 67.72 | |
| QimeraBackbone=ResNet-50, Bit-width=4 bits2026.06 | 66.25 | |
| SQuantBackbone=ResNet-18, Bit-width=4 bits2026.06 | 66.14 | |
| EmaQ-LTBackbone=ResNet-18, Bit-width=4 bits2026.06 | 66.14 | |
| EmaQ-LTBackbone=ResNet-50, Bit-width=2 bits2026.06 | 65.68 | |
| ZAQBackbone=ResNet-50, Bit-width=2 bits2026.06 | 65.52 | |
| ARC+AITBackbone=ResNet-18, Bit-width=4 bits2026.06 | 65.43 | |
| QimeraBackbone=ResNet-18, Bit-width=4 bits2026.06 | 63.84 | |
| ZeroQBackbone=ResNet-50, Bit-width=2 bits2026.06 | 63.12 | |
| Choi et al.Backbone=ResNet-50, Bit-width=2 bits2026.06 | 63 | |
| BatchQuantBackbone=ResNet-50, Bit-width=2 bits2026.06 | 62.21 | |
| EmaQ-LTBackbone=ResNet-18, Bit-width=2 bits2026.06 | 61.13 | |
| ZeroQBackbone=ResNet-18, Bit-width=4 bits2026.06 | 26 |