MoFME, an adverse weather removal AI jointly developed under the framework of BAIR Open Research Commons, was accepted to AAAI2024
Panasonic Holdings Co., Ltd. (Panasonic HD) has jointly developed an Adverse weather removal AI with researchers from the University of California, Berkeley (UC Berkeley), Nanjing University, and Peking University that improves image recognition accuracy by removing rain, snow, fog, and other elements from images that can significantly reduce image recognition accuracy. This method demonstrated image restoration performance that improved recognition accuracy over conventional methods while saving over 72% of parameters and 39% of inference time in image recognition and segmentation tasks for multiple adverse weather images.
- Accepted Paper
"Efficient Deweather Mixture-of-Experts with Uncertainty-aware Feature-wise Linear Modulation"
Rongyu Zhang, Yulin Luo, Jiaming Liu, Huanrui Yang, Zhen Dong, Denis Gudovskiy, Tomoyuki Okuno, Yohei Nakata, Kurt Keutzer, Yuan Du, Shanghang Zhang
https://arxiv.org/abs/2312.16610
Outdoor image recognition AI applications are becoming increasingly common in fields such as mobility and infrastructure.
By representing different weather parameters as weights, the joint development team created a technology that can reliably remove the influence of adverse weather using a small number of parameters in a single model that can handle multiple weather elements and tasks. This proposed method could be used in a variety of situations where high-precision image recognition is required in all weather conditions, such as danger detection systems for in-vehicle sensors and security cameras.
The technology has been internationally recognized and accepted into the 38th Annual AAAI Conference on Artificial Intelligence (AAAI 2024), a top conference for AI and machine learning technology, and will be presented at the conference, to be held in Vancouver, Canada from February 22, 2024 to February 25, 2024.
Please also refer to the press release.
- Panasonic News Room
[Press Release] Panasonic HD Develops Adverse Weather Removal AI That Removes Rain, Snow, Fog, Other Elements That Degrade Image Recognition Accuracy The technology uses one-third the number of parameters compared to previous
https://news.panasonic.com/global/press/en240216-2
