Autoencoder-based background reconstruction and foreground segmentation with background noise estimation
This addresses the open problem of background subtraction in computer vision, particularly for videos with camera movement, offering an unsupervised method that outperforms existing approaches.
The paper tackles dynamic scene background reconstruction and foreground segmentation by modeling the background as a low-dimensional manifold using an autoencoder that also predicts background noise, enabling pixel-dependent thresholds for segmentation. It achieves state-of-the-art results on CDnet 2014 and LASIESTA datasets, with significant improvements in moving camera scenarios.
Even after decades of research, dynamic scene background reconstruction and foreground object segmentation are still considered as open problems due various challenges such as illumination changes, camera movements, or background noise caused by air turbulence or moving trees. We propose in this paper to model the background of a frame sequence as a low dimensional manifold using an autoencoder and compare the reconstructed background provided by this autoencoder with the original image to compute the foreground/background segmentation masks. The main novelty of the proposed model is that the autoencoder is also trained to predict the background noise, which allows to compute for each frame a pixel-dependent threshold to perform the foreground segmentation. Although the proposed model does not use any temporal or motion information, it exceeds the state of the art for unsupervised background subtraction on the CDnet 2014 and LASIESTA datasets, with a significant improvement on videos where the camera is moving. It is also able to perform background reconstruction on some non-video image datasets.