Domain-Specific Suppression for Adaptive Object Detection

Wang, Yu; Zhang, Rui; Zhang, Shuo; Li, Miao; Xia, Yangyang; Zhang, XiShan; Liu, Shaoli

doi:10.1109/cvpr46437.2021.00948

Cited by 65 publications

(17 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…α and β are set 0.1 and 1, respectively. We set an extremely low threshold (0.05) to accelerate node sampling to establish (Chen et al 2018) 34.3 EPM (Hsu et al 2020a) 49.0 +9.2 DSS (Wang et al 2021) 44.5 +9.8 MEGA-CDA (VS et al 2021) 44.8 +10.5 RPNPA (Zhang, Wang, and Mao 2021) 45.7 +11.1 UMT (Deng et al 2021) 43.1 +8.8 Source Only (Hsu et al 2020a) 39.8 Baseline (Hsu et al 2020a) 45.9…”

Section: Implementation Detailsmentioning

confidence: 99%

SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation

Liu

Yao

et al. 2022

AAAI

View full text Add to dashboard Cite

The domain gap severely limits the transferability and scalability of object detectors trained in a specific domain when applied to a novel one. Most existing works bridge the domain gap by minimizing the domain discrepancy in the category space and aligning category-agnostic global features. Though great success, these methods model domain discrepancy with prototypes within a batch, yielding a biased estimation of domain-level distribution. Besides, the category-agnostic alignment leads to the disagreement of class-specific distributions in the two domains, further causing inevitable classification errors. To overcome these two challenges, we propose a novel Semantic Conditioned AdaptatioN (SCAN) framework such that well-modeled unbiased semantics can support semantic conditioned adaptation for precise domain adaptive object detection. Specifically, class-specific semantics crossing different images in the source domain are graphically aggregated as the input to learn an unbiased semantic paradigm incrementally. The paradigm is then sent to a lightweight manifestation module to obtain conditional kernels to serve as the role of extracting semantics from the target domain for better adaptation. Subsequently, conditional kernels are integrated into global alignment to support the class-specific adaptation in a well-designed Conditional Kernel guided Alignment (CKA) module. Meanwhile, rich knowledge of the unbiased paradigm is transferred to the target domain with a novel Graph-based Semantic Transfer (GST) mechanism, yielding the adaptation in the category-based feature space. Comprehensive experiments conducted on three adaptation benchmarks demonstrate that SCAN outperforms existing works by a large margin.

show abstract

Section: Implementation Detailsmentioning

confidence: 99%

SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation

Liu

Yao

et al. 2022

AAAI

View full text Add to dashboard Cite

show abstract

“…In contrast, in this paper, the model is updated in a self-supervised way to improve the detection performance. There have also been various works in segmentation and detection that adapt the network [10,46,48,51,54]. However, they have access to only a fixed set of images and there is no mechanism to interact with an environment.…”

Section: Related Workmentioning

confidence: 99%

“…We believe there are strong self-supervised signals in the inference phase that an embodied agent can leverage via interacting with its environment to adapt the model. There has been work to adapt object detection models in an unsupervised way (e.g., [10,46,51,54]). However, they assume a pre-recorded set of observations during inference.…”

Section: Introductionmentioning

confidence: 99%

Interactron: Embodied Adaptive Object Detection

Kotar¹,

Mottaghi²

2022

Preprint

View full text Add to dashboard Cite

Over the years various methods have been proposed for the problem of object detection. Recently, we have witnessed great strides in this domain owing to the emergence of powerful deep neural networks. However, there are typically two main assumptions common among these approaches. First, the model is trained on a fixed training set and is evaluated on a pre-recorded test set. Second, the model is kept frozen after the training phase, so no further updates are performed after the training is finished. These two assumptions limit the applicability of these methods to real-world settings. In this paper, we propose Interactron, a method for adaptive object detection in an interactive setting, where the goal is to perform object detection in images observed by an embodied agent navigating in different environments. Our idea is to continue training during inference and adapt the model at test time without any explicit supervision via interacting with the environment. Our adaptive object detection model provides a 11.8 point improvement in AP (and 19.1 points in AP 50 ) over DETR [5], a recent, highperformance object detector. Moreover, we show that our object detection model adapts to environments with completely different appearance characteristics, and its performance is on par with a model trained with full supervision for those environments. The code will be released soon.

show abstract

“…Deng et al [8] adopted the CycleGAN and Mean Teacher model to eliminate the model bias. For the feature disentanglement, there are some methods [25,47,50] extracting the domain-invariant features for domain shift. Although these methods have been demonstrated to be effective, we proposed a new observation with domain specific augmentation consistency for DAOD.…”

Section: Related Workmentioning

confidence: 99%

Frequency Spectrum Augmentation Consistency for Domain Adaptive Object Detection

Li¹,

Han²,

Wang³

2021

Preprint

View full text Add to dashboard Cite

Domain adaptive object detection (DAOD) aims to improve the generalization ability of detectors when the training and test data are from different domains. Considering the significant domain gap, some typical methods, e.g., CycleGAN-based methods, adopt the intermediate domain to bridge the source and target domains progressively. However, the CycleGAN-based intermediate domain lacks the pix-or instance-level supervision for object detection, which leads to semantic differences. To address this problem, in this paper, we introduce a Frequency Spectrum Augmentation Consistency (FSAC) framework with four different low-frequency filter operations. In this way, we can obtain a series of augmented data as the intermediate domain.Concretely, we propose a two-stage optimization framework. In the first stage, we utilize all the original and augmented source data to train an object detector. In the second stage, augmented source and target data with pseudo labels are adopted to perform the self-training for prediction consistency. And a teacher model optimized using Mean Teacher is used to further revise the pseudo labels. In the experiment, we evaluate our method on the single-and compound-target DAOD separately, which demonstrate the effectiveness of our method. And our code is avaliable at https://github.com/ruiliu-code/FSAC.

show abstract

Domain-Specific Suppression for Adaptive Object Detection

Cited by 65 publications

References 27 publications

SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation

SCAN: Cross Domain Object Detection with Semantic Conditioned Adaptation

Interactron: Embodied Adaptive Object Detection

Frequency Spectrum Augmentation Consistency for Domain Adaptive Object Detection

Contact Info

Product

Resources

About