Real-Time Quantized Image Super-Resolution on Mobile NPUs, Mobile AI 2021 Challenge: Report

Abstract

Image super-resolution is one of the most popular computer vision problems with many important applications to mobile devices. While many solutions have been proposed for this task, they are usually not optimized even for common smartphone AI hardware, not to mention more constrained smart TV platforms that are often supporting INT8 inference only. To address this problem, we introduce the first Mobile AI challenge, where the target is to develop an end-to-end deep learning-based image super-resolution solutions that can demonstrate a real-time performance on mobile or edge NPUs. For this, the participants were provided with the DIV2K dataset and trained quantized models to do an efficient 3X image upscaling. The runtime of all models was evaluated on the Synaptics VS680 Smart Home board with a dedicated NPU capable of accelerating quantized neural networks. The proposed solutions are fully compatible with all major mobile AI accelerators and are capable of reconstructing Full HD images under 40-60 ms while achieving high fidelity results. A detailed description of all models developed in the challenge is provided in this paper.

Five Forces Model of Computational Power: A Comprehensive Measure Method

“…The computational power of a data center is determined by data processing capability, data storage capability and data circulation capability. In the process of responding to the industrial trend of the new generation of digital technology represented by big data and AI [20], data processing capability can be divided into general computational power represented by CPU and intelligent computational power represented by GPU and AI chip. The former is mainly used to perform general tasks, while the latter mainly undertakes computationally intensive tasks such as graphics display, big data analysis, signal processing, artificial intelligence and physical simulation.…”