SOTAVerified

Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation

2019-01-10CVPR 2019Code Available0· sign in to hype

Chenxi Liu, Liang-Chieh Chen, Florian Schroff, Hartwig Adam, Wei Hua, Alan Yuille, Li Fei-Fei

Code Available — Be the first to reproduce this paper.

Reproduce

Code

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.

Tasks

Benchmark Results

DatasetModelMetricClaimedVerifiedStatus
ADE20KAuto-DeepLab-LValidation mIoU43.98Unverified
ADE20K valAuto-DeepLab-LmIoU43.98Unverified
Cityscapes testAuto-DeepLab-LMean IoU (class)82.1Unverified
Cityscapes valAuto-DeepLab-LmIoU80.33Unverified
PASCAL VOC 2012 testAuto-DeepLab-LMean IoU85.6Unverified
PASCAL VOC 2012 valAuto-DeepLab-LmIoU82.04Unverified

Reproductions