Adversarial Branch Architecture Search for Unsupervised Domain Adaptation

Abstract

Unsupervised Domain Adaptation (UDA) is a key field in visual recognition, as it enables robust performances across different visual domains. In the deep learning era, the performance of UDA methods has been driven by better losses and by improved network architectures, specifically the addition of auxiliary domain-alignment branches to pre-trained backbones. However, all the neural architectures proposed so far are hand-crafted, which might hinder further progress.

On Evolving Attention Towards Domain Adaptation

“…Different from the existing literature, in this paper, we investigate the possibility and an effective solution to search for optimal architectures towards better domain adaptation. Concurrently, we find that Li et al [26] and Robbiano et al [40] have investigated a similar topic: the generalization abilities of architectures cross domains. The differences are two-fold: 1) we propose a novel search space for diverse attention configurations, which is different from the search space of AdaptNAS [26] (akin to NASNet [57]) and ABAS [40] (just change the architecture of the auxiliary adversarial branch); 2) we focus on an effective NAS protocol to search for optimal attention configuration towards UDA tasks.…”

“…Differences from Concurrent Methods. Concurrently, Liet al [26] and Robbianoet al [40] also propose to seek for better transferable network architectures. The differences between our EvoADA and them are: 1) on the design of search space: AdaptNAS [26] adopted the search space of NASNet [57] that includes basic operations in CNNs (e.g., different pooling or convolution operations) and seeks for optimal structures in a cell perspective, ABAS [40] only changes the structure of the auxiliary branch.…”

“…Concurrently, Liet al [26] and Robbianoet al [40] also propose to seek for better transferable network architectures. The differences between our EvoADA and them are: 1) on the design of search space: AdaptNAS [26] adopted the search space of NASNet [57] that includes basic operations in CNNs (e.g., different pooling or convolution operations) and seeks for optimal structures in a cell perspective, ABAS [40] only changes the structure of the auxiliary branch. In EvoADA, we design a new space to produce diverse attention configurations and apply different modules on various intermediate layers of the backbone network, which is a more generalized manner; 2) on the architecture search process: AdaptNAS adopts a gradient-based differentiable scheme [31], which might result in sub-optimal solutions; ABAS leverages the BOHB [14], a Bayesian-based hyperparameter optimization method, which is not suitable for a high-dimensional optimization problem.…”

“…Baselines. We compare with multiple state-of-the-art approaches: DANN [15], JAN [33], OSBP [44], CDAN [32], IBN-Net [34], MCD [43], SAN [4], TADA [50], BSP [7], SAFN [53], STA [30], CADA-A [24], ETN [5], CRST [58], BA 3 US [28], DCAN [25], MMT [17], PAN [47], SHOT [27], and ABAS [40]. Among them, TADA [50], CADA-A [24], DCAN [25] are the competitive approaches [43] 88.6 98.5 100.0 92.2 69.5 69.7 86.5 CRST [58] 89.4 98.9 100.0 88.7 72.6 70.9 86.8 CDAN [32] 94.1 98.6 100.0 92.9 71.0 69.3 87.7 TADA [50] 94.3 98.7 99.8 91.6 72.9 73.0 88.4 BSP [7] 93.3 98.2 100.0 93.0 73.6 72.6 88.5 CADA-A [24] 96.8 99.0 99.8 93.4 71.7 70.5 88.5 SHOT [27] 90.9 98.8 99.9 93.1 74.5 74.8 88.7 SHOT+Ours 94.0 97.9 100.0 94.2 74.6 74.9 89.3…”

Domain Adaptation by Maximizing Population Correlation with Neural Architecture Search

“…Li and Peng (2020) propose a DARTS-like method for DA, which combines DARTS and DA into one framework. Robbiano et al (2021) aim to learn a auxiliary branch network from data for an adversarial DA method. In this paper, different from those works, we aim to leverage NAS to search optimal neural architectures for the proposed DAMPC method.…”