Biased Mixtures Of Experts: Enabling Computer Vision Inference Under Data Transfer Limitations

TL;DR

This paper introduces biased mixtures of experts for computer vision, optimizing model inference under data transfer constraints by selectively activating experts based on data availability, leading to improved performance.

Contribution

It presents a novel mixture-of-experts framework that biases expert selection to adapt to data transfer limitations, formulated as a convex optimization problem.

Findings

Outperforms previous methods under data transfer constraints

Effective across detection, super-resolution, and video classification

Biasing experts improves accuracy with limited data

Abstract

We propose a novel mixture-of-experts class to optimize computer vision models in accordance with data transfer limitations at test time. Our approach postulates that the minimum acceptable amount of data allowing for highly-accurate results can vary for different input space partitions. Therefore, we consider mixtures where experts require different amounts of data, and train a sparse gating function to divide the input space for each expert. By appropriate hyperparameter selection, our approach is able to bias mixtures of experts towards selecting specific experts over others. In this way, we show that the data transfer optimization between visual sensing and processing can be solved as a convex optimization problem.To demonstrate the relation between data availability and performance, we evaluate biased mixtures on a range of mainstream computer vision problems, namely: (i) single shot detection, (ii) image super resolution, and (iii) realtime video action classification. For all cases, and when experts constitute modified baselines to meet different limits on allowed data utility, biased mixtures significantly outperform previous work optimized to meet the same constraints on available data.