%0 Journal Article %1 journals/corr/abs-1901-02985 %A Liu, Chenxi %A Chen, Liang-Chieh %A Schroff, Florian %A Adam, Hartwig %A Hua, Wei %A Yuille, Alan L. %A Fei-Fei, Li %D 2019 %J Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) %K from:fbuckermann reserved thema thema:neural_architecture_search %T Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation. %U https://openaccess.thecvf.com/content_CVPR_2019/html/Liu_Auto-DeepLab_Hierarchical_Neural_Architecture_Search_for_Semantic_Image_Segmentation_CVPR_2019_paper.html %X Recently, Neural Architecture Search (NAS) has successfully identified neural network architectures that exceed human designed ones on large-scale image classification. In this paper, we study NAS for semantic image segmentation. Existing works often focus on searching the repeatable cell structure, while hand-designing the outer network structure that controls the spatial resolution changes. This choice simplifies the search space, but becomes increasingly problematic for dense image prediction which exhibits a lot more network level architectural variations. Therefore, we propose to search the network level structure in addition to the cell level structure, which forms a hierarchical architecture search space. We present a network level search space that includes many popular designs, and develop a formulation that allows efficient gradient-based architecture search (3 P100 GPU days on Cityscapes images). We demonstrate the effectiveness of the proposed method on the challenging Cityscapes, PASCAL VOC 2012, and ADE20K datasets. Auto-DeepLab, our architecture searched specifically for semantic image segmentation, attains state-of-the-art performance without any ImageNet pretraining.

@article{journals/corr/abs-1901-02985, abstract = {Recently, Neural Architecture Search (NAS) has successfully identified neural network architectures that exceed human designed ones on large-scale image classification. In this paper, we study NAS for semantic image segmentation. Existing works often focus on searching the repeatable cell structure, while hand-designing the outer network structure that controls the spatial resolution changes. This choice simplifies the search space, but becomes increasingly problematic for dense image prediction which exhibits a lot more network level architectural variations. Therefore, we propose to search the network level structure in addition to the cell level structure, which forms a hierarchical architecture search space. We present a network level search space that includes many popular designs, and develop a formulation that allows efficient gradient-based architecture search (3 P100 GPU days on Cityscapes images). We demonstrate the effectiveness of the proposed method on the challenging Cityscapes, PASCAL VOC 2012, and ADE20K datasets. Auto-DeepLab, our architecture searched specifically for semantic image segmentation, attains state-of-the-art performance without any ImageNet pretraining.}, added-at = {2020-10-19T15:36:19.000+0200}, author = {Liu, Chenxi and Chen, Liang-Chieh and Schroff, Florian and Adam, Hartwig and Hua, Wei and Yuille, Alan L. and Fei-Fei, Li}, biburl = {https://www.bibsonomy.org/bibtex/2c1e71c6f709c085d560409e5a63491b9/uw_ws20_ml}, ee = {http://arxiv.org/abs/1901.02985}, interhash = {12032b2793bab10d8199c64da666cd3c}, intrahash = {c1e71c6f709c085d560409e5a63491b9}, journal = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, keywords = {from:fbuckermann reserved thema thema:neural_architecture_search}, month = {june}, timestamp = {2020-11-26T14:23:34.000+0100}, title = {Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation.}, url = {https://openaccess.thecvf.com/content_CVPR_2019/html/Liu_Auto-DeepLab_Hierarchical_Neural_Architecture_Search_for_Semantic_Image_Segmentation_CVPR_2019_paper.html}, year = 2019 }