Is In-Domain Data Really Needed? A Pilot Study on Cross-Domain Calibration for Network Quantization

by   Haichao Yu, et al.

Post-training quantization methods use a set of calibration data to compute quantization ranges for network parameters and activations. The calibration data usually comes from the training dataset which could be inaccessible due to sensitivity of the data. In this work, we want to study such a problem: can we use out-of-domain data to calibrate the trained networks without knowledge of the original dataset? Specifically, we go beyond the domain of natural images to include drastically different domains such as X-ray images, satellite images and ultrasound images. We find cross-domain calibration leads to surprisingly stable performance of quantized models on 10 tasks in different image domains with 13 different calibration datasets. We also find that the performance of quantized models is correlated with the similarity of the Gram matrices between the source and calibration domains, which can be used as a criterion to choose calibration set for better performance. We believe our research opens the door to borrow cross-domain knowledge for network quantization and compression.



There are no comments yet.


page 4

page 5

page 7


An Underexplored Dilemma between Confidence and Calibration in Quantized Neural Networks

Modern convolutional neural networks (CNNs) are known to be overconfiden...

Towards Semantic Big Graph Analytics for Cross-Domain Knowledge Discovery

In recent years, the size of big linked data has grown rapidly and this ...

Fast Adjustable Threshold For Uniform Neural Network Quantization

Neural network quantization procedure is the necessary step for porting ...

On Calibration and Out-of-domain Generalization

Out-of-domain (OOD) generalization is a significant challenge for machin...

On the limits of cross-domain generalization in automated X-ray prediction

This large scale study focuses on quantifying what X-rays diagnostic pre...

Ranking Distance Calibration for Cross-Domain Few-Shot Learning

Recent progress in few-shot learning promotes a more realistic cross-dom...

Fine-grained Data Distribution Alignment for Post-Training Quantization

While post-training quantization receives popularity mostly due to its e...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.