Toward Multi-Domain and Long-Tailed Quantization via Feature Alignment and Scaling
About
Quantizing deep neural networks is essential for efficient inference on resource-constrained devices. However, most existing methods are designed for single-domain and class-balanced data, leaving practical settings with domain shifts or severe class imbalance underexplored. We address these challenges with Efficient Multi-Domain Alignment Quantization (EmaQ), which aligns domain distributions through a CDF-based projection and uses sensitivity-aware weight aggregation to stabilize multi-domain quantization. We further extend EmaQ to EmaQ-LT for long-tailed quantization by introducing class-conditioned variance scaling and confidence-based logit adjustment to mitigate majority-class overconfidence. Theoretical analyses establish convergence guarantees and motivate the proposed sensitivity and scaling mechanisms. Experiments on standard, multi-domain (Office-31, Digits), and long-tailed (SynDigits-LT, CIFAR-10-LT, CIFAR-100-LT) benchmarks show that EmaQ and EmaQ-LT achieve strong low-bit performance under domain shift and class imbalance.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Image Classification | Office-31 | Average Accuracy80.6 | 357 | |
| Image Classification | CIFAR-100 LT | Top-1 Acc55.62 | 225 | |
| Image Classification | SVHN | Top-1 Accuracy96.3 | 209 | |
| Image Classification | CIFAR-10-LT gamma=50 (test) | Accuracy74.4 | 68 | |
| Domain Adaptation | MNIST to MNIST-M (test) | Acc (Target)96.1 | 47 | |
| Image Classification | Digits | Accuracy (MNIST-M Source)98.78 | 41 | |
| Domain Adaptation | MNIST to SVHN (test) | Accuracy59.9 | 34 | |
| Image Classification | CIFAR-10-LT gamma=10 (test) | Accuracy81.36 | 34 | |
| Image Classification | SynDigits-LT gamma=10 | Accuracy96.45 | 32 | |
| Image Classification | SynDigits-LT gamma=50 | Accuracy89.8 | 31 |