Contrastive Regression for Domain Adaptation on Gaze Estimation

Wang, Yaoming; Jiang, Yangzhou; Li, Jin; Ni, Bingbing; Dai, Wenrui; Li, Chenglin; Xiong, Hongkai; Li, Teng

doi:10.1109/cvpr52688.2022.01877

Cited by 44 publications

(25 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another work by Wang et al (2022) proposes to improve domain adaptation for gaze estimation by adding a contrastive loss term to the L1 loss. They show that their approach is beneficial for adapting a gaze estimation model from one dataset (i.e., one domain) to another, but the approach produces no benefits and even reduces performance for the source dataset.…”

Section: Related Workmentioning

confidence: 99%

Supervised Contrastive Regression

Zha¹,

Cao²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Deep regression models typically learn in an end-to-end fashion and do not explicitly try to learn a regression-aware representation. Their representations tend to be fragmented and fail to capture the continuous nature of regression tasks. In this paper, we propose Supervised Contrastive Regression (SupCR), a framework that learns a regression-aware representation by contrasting samples against each other based on their target distance. SupCR is orthogonal to existing regression models, and can be used in combination with such models to improve performance. Extensive experiments using five real-world regression datasets that span computer vision, human-computer interaction, and healthcare show that using SupCR achieves the state-of-the-art performance and consistently improves prior regression baselines on all datasets, tasks, and input modalities. SupCR also improves robustness to data corruptions, resilience to reduced training data, performance on transfer learning, and generalization to unseen targets.

show abstract

Section: Related Workmentioning

confidence: 99%

Supervised Contrastive Regression

Zha¹,

Cao²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The most notable version of contrastive loss is NT-Xent, highlighted by Chen et al [9]. Previously contrastive loss has been used in the context of classification tasks and is thus not best suited for regression tasks [10]. Wang et al [10] proposed a contrastive loss by using the distribution of the labels as the weights with KL-divergence.…”

Section: Contrastive Learningmentioning

confidence: 99%

“…Previously contrastive loss has been used in the context of classification tasks and is thus not best suited for regression tasks [10]. Wang et al [10] proposed a contrastive loss by using the distribution of the labels as the weights with KL-divergence. Jindal et al [20] recently proposed a GazeCRL, a contrastive learning framework for gaze estimation.…”

Section: Contrastive Learningmentioning

confidence: 99%

“…We propose CRWL, a novel contrastive regression loss function for gaze estimation. Contrary to classification tasks, the distancing method has to be scaled relative to the sample's proximity to the anchor sample inferred by the labels [10]. Also it should be scaled accordingly since the similarities between samples are relatively similar between gaze samples.…”

Section: Contrastive Regression Loss For Gaze Estimationmentioning

confidence: 99%

“…Contrastive learning has been used in many SSL training frameworks. However, the general contrastive loss is most suited for classification tasks [10]. With regression, the slightest changes in the latent vector representation of features could hinder the performance drastically.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Contrastive Weighted Learning for Near-Infrared Gaze Estimation

Lee¹

2022

Preprint

View full text Add to dashboard Cite

Appearance-based gaze estimation has been very successful with the use of deep learning. Many following works improved domain generalization for gaze estimation. However, even though there has been much progress in domain generalization for gaze estimation, most of the recent work have been focused on cross-dataset performance-accounting for different distributions in illuminations, head pose, and lighting. Although improving gaze estimation in different distributions of RGB images is important, near-infrared image based gaze estimation is also critical for gaze estimation in dark settings. Also there are inherent limitations relying solely on supervised learning for regression tasks. This paper contributes to solving these problems and proposes GazeCWL, a novel framework for gaze estimation with near-infrared images using contrastive learning. This leverages adversarial attack techniques for data augmentation and a novel contrastive loss function specifically for regression tasks that effectively clusters the features of different samples in the latent space. Our model outperforms previous domain generalization models in infrared image based gaze estimation and outperforms the baseline by 45.6% while improving the state-of-the-art by 8.6%, we demonstrate the efficacy of our method.

show abstract

Towards an Accurate 3D Deformable Eye Model for Gaze Estimation

Kuang

Kephart

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Contrastive Regression for Domain Adaptation on Gaze Estimation

Cited by 44 publications

References 25 publications

Supervised Contrastive Regression

Supervised Contrastive Regression

Contrastive Weighted Learning for Near-Infrared Gaze Estimation

Towards an Accurate 3D Deformable Eye Model for Gaze Estimation

Contact Info

Product

Resources

About