| dc.description.abstract | The availability of dehazing datasets has enabled various deep learning techniques to perform effectively on hazy images. Most of the developed frameworks focus on removing homogeneous haze. However, homogeneous centric methods produce suboptimal results on nonhomogeneous haze. The primary reason is that the architectures devised to handle homogeneous haze fail to address the non uniformity of haze in non homogeneous case. The secondary reaon is the unavailability of enough data for the non homogeneous scenario. Al though many works cite the lack of data as a primary concern for poor performance, we find that even if the homogeneous-centric networks are trained with non-homogeneous data, the produced results are sub standard. Hence, there is a requirement for a network architecture that can handle non homogeneous haze in a better way. In this work, we propose to use multiple attention mechanisms in parallel along with pretrained ConvNeXt blocks. Specifically, we use pixel, channel, and residual channel attention mechanisms. Pixel attention can complement the channel attention in dealing with spacevariant haze when connected in parallel. On the other hand, residual channel attention fetches hazy imagerelated features and caters to better information flow towards the output. Concatenating the attention based features can yield better results as compared to the existing approaches. | |
