A Mixed Quantization Network for Computationally Efficient Mobile Inverse Tone Mapping

TL;DR

This paper introduces a mixed quantization neural network designed for efficient inverse tone mapping, enabling high-quality HDR image recovery on mobile devices with significantly reduced latency and memory usage.

Contribution

The paper presents a novel mixed quantization scheme combined with efficient neural network operations for real-time inverse tone mapping on resource-constrained devices.

Findings

Achieves comparable performance to state-of-the-art methods on benchmarks.

Provides up to 10x faster latency and 25x less memory consumption.

Demonstrates effectiveness of attention mechanisms and quantization schemes in ablation studies.

Abstract

Recovering a high dynamic range (HDR) image from a single low dynamic range (LDR) image, namely inverse tone mapping (ITM), is challenging due to the lack of information in over- and under-exposed regions. Current methods focus exclusively on training high-performing but computationally inefficient ITM models, which in turn hinder deployment of the ITM models in resource-constrained environments with limited computing power such as edge and mobile device applications. To this end, we propose combining efficient operations of deep neural networks with a novel mixed quantization scheme to construct a well-performing but computationally efficient mixed quantization network (MQN) which can perform single image ITM on mobile platforms. In the ablation studies, we explore the effect of using different attention mechanisms, quantization schemes, and loss functions on the performance of MQN in ITM tasks. In the comparative analyses, ITM models trained using MQN perform on par with the state-of-the-art methods on benchmark datasets. MQN models provide up to 10 times improvement on latency and 25 times improvement on memory consumption.