Human Pose Estimation in Extremely Low-Light Conditions
This addresses a practical problem for computer vision applications in low-light environments, but it is incremental as it builds on existing pose estimation methods with a new dataset and training strategy.
The paper tackles human pose estimation in extremely low-light conditions by developing a new dataset with aligned well-lit images and a model that uses this privileged information, achieving outstanding performance on real low-light images.
We study human pose estimation in extremely low-light images. This task is challenging due to the difficulty of collecting real low-light images with accurate labels, and severely corrupted inputs that degrade prediction quality significantly. To address the first issue, we develop a dedicated camera system and build a new dataset of real low-light images with accurate pose labels. Thanks to our camera system, each low-light image in our dataset is coupled with an aligned well-lit image, which enables accurate pose labeling and is used as privileged information during training. We also propose a new model and a new training strategy that fully exploit the privileged information to learn representation insensitive to lighting conditions. Our method demonstrates outstanding performance on real extremely low light images, and extensive analyses validate that both of our model and dataset contribute to the success.