Fast and Accurate Camera Scene Detection on Smartphones

TL;DR

This paper introduces a new dataset and a highly efficient CNN model for accurate camera scene detection on smartphones, achieving near-perfect accuracy and real-time performance on modern mobile hardware.

Contribution

It defines the camera scene detection problem, creates the first comprehensive dataset (CamSDD), and proposes a NPU-friendly CNN model that balances accuracy and speed.

Findings

Top-3 accuracy of 99.5% on CamSDD

Over 200 FPS on recent mobile SoCs

Effective real-world performance analysis

Abstract

AI-powered automatic camera scene detection mode is nowadays available in nearly any modern smartphone, though the problem of accurate scene prediction has not yet been addressed by the research community. This paper for the first time carefully defines this problem and proposes a novel Camera Scene Detection Dataset (CamSDD) containing more than 11K manually crawled images belonging to 30 different scene categories. We propose an efficient and NPU-friendly CNN model for this task that demonstrates a top-3 accuracy of 99.5% on this dataset and achieves more than 200 FPS on the recent mobile SoCs. An additional in-the-wild evaluation of the obtained solution is performed to analyze its performance and limitation in the real-world scenarios. The dataset and pre-trained models used in this paper are available on the project website.