Contrastive Domain Adaptation

Thota, Mamatha; Leontidis, Georgios

doi:10.48550/arxiv.2103.15566

Cited by 4 publications

(4 citation statements)

References 37 publications

(61 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, contrastive learning in the context of DA remains underexplored. In Thota and Leontidis (2021), the authors propose to extend the contrastive learning approach to a situation where none of the data domains contain any labeled data. In our work, we combine ideas of contrastive learning on unlabeled samples and supervised learning using the labeled source domain data to build our UDA method for astronomy.…”

Section: Methodsmentioning

confidence: 99%

DeepAstroUDA: semi-supervised universal domain adaptation for cross-survey galaxy morphology classification and anomaly detection

Ćiprijanović

Lewis

Pedro

et al. 2023

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

Artificial intelligence methods show great promise in increasing the quality and speed of work with large astronomical datasets, but the high complexity of these methods leads to the extraction of dataset-specific, non-robust features. Therefore, such methods do not generalize well across multiple datasets. We present a universal domain adaptation method, DeepAstroUDA, as an approach to overcome this challenge.This algorithm performs semi-supervised domain adaptation and can be applied to datasets with different data distributions and class overlaps. Non-overlapping classes can be present in any of the two datasets (the labeled source domain, or the unlabeled target domain), and the method can even be used in the presence of unknown classes. We apply our method to three examples of galaxy morphology classification tasks of different complexities (3-class and 10-class problems), with anomaly detection: 1) datasets created after different numbers of observing years from a single survey (LSST mock data of 1 and 10 years of observations); 2) data from different surveys (SDSS and DECaLS); and 3) data from observing fields with different depths within one survey (wide field and Stripe 82 deep field of SDSS). For the first time, we demonstrate the successful use of domain adaptation between very discrepant observational datasets. DeepAstroUDA is capable of bridging the gap between two astronomical surveys, increasing classification accuracy in both domains (up to 40% on the unlabeled data), and making model performance consistent across datasets. Furthermore, our method also performs well as an anomaly detection algorithm and successfully clusters unknown class samples even in the unlabeled target dataset.

show abstract

Section: Methodsmentioning

confidence: 99%

DeepAstroUDA: semi-supervised universal domain adaptation for cross-survey galaxy morphology classification and anomaly detection

Ćiprijanović

Lewis

Pedro

et al. 2023

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…However, contrastive learning in the context of domain adaptation remains underexplored. In Thota and Leontidis (2021), authors propose to extend contrastive learning approach to a situation where none of the data domains contain any labeled data. In our work we combine ideas of contrastive learning on unlabeled samples and supervised learning using the labeled source domain data to build our UDA method for astronomy.…”

Section: Methodsmentioning

confidence: 99%

DeepAstroUDA: Semi-Supervised Universal Domain Adaptation for Cross-Survey Galaxy Morphology Classification and Anomaly Detection

Ćiprijanović¹,

Lewis²,

Pedro³

et al. 2023

Preprint

View full text Add to dashboard Cite

Artificial intelligence methods show great promise in increasing the quality and speed of work with large astronomical datasets, but the high complexity of these methods leads to the extraction of dataset-specific, nonrobust features. Therefore, such methods do not generalize well across multiple datasets. We present a universal domain adaptation method, DeepAstroUDA, as an approach to overcome this challenge. This algorithm performs semi-supervised domain adaptation and can be applied to datasets with different data distributions and class overlaps. Non-overlapping classes can be present in any of the two datasets (the labeled source domain, or the unlabeled target domain), and the method can even be used in the presence of unknown classes. We apply our method to three examples of galaxy morphology classification tasks of different complexities (3-class and 10-class problems), with anomaly detection: 1) datasets created after different numbers of observing years from a single survey (LSST mock data of 1 and 10 years of observations); 2) data from different surveys (SDSS and DECaLS); and 3) data from observing fields with different depths within one survey (wide field and Stripe 82 deep field of SDSS). For the first time, we demonstrate the successful use of domain adaptation between very discrepant observational datasets. DeepAstroUDA is capable of bridging the gap between two astronomical surveys, increasing classification accuracy in both domains (up to 40% on the unlabeled data), and making model performance consistent across datasets. Furthermore, our method also performs well as an anomaly detection algorithm and successfully clusters unknown class samples even in the unlabeled target dataset.

show abstract

“…To address the sparse reward problem, several expert trajectories are incorporated into the replay buffer. However, CURL, and more generally, selfcontrastive representation learning methods, exhibit limited performance in domain adaptation, particularly in bridging the Sim2Real gap [27]. To overcome these challenges, we combine CURL with segmentation-driven domain shifts to enhance its Sim2Real transferability.…”

Section: B Rl Based Visual Controlmentioning

confidence: 99%

Seg-CURL: Segmented Contrastive Unsupervised Reinforcement Learning for Sim-to-Real in Visual Robotic Manipulation

Hassan

Hussain

2023

IEEE Access

View full text Add to dashboard Cite

Training image-based reinforcement learning (RL) agents are sample-inefficient, limiting their effectiveness in real-world manipulation tasks. Sim2Real, which involves training in simulations and transferring to the real world, effectively reduces the dependence on real data. However, the performance of the transferred agent degrades due to the visual difference between the two environments. This research presents a low-cost segmentation-driven unsupervised RL framework (Seg-CURL) to solve the Sim2Real problem. We transform the input RGB views to the proposed semantic segmentation-based canonical domain. Our method incorporates two levels of Sim2Real: task-level Sim2Real, which transfers the RL agent to the real world, and observation-level Sim2Real, which transfers the simulated U-nets to segment real-world scenes. Specifically, we first train contrastive unsupervised RL(CURL) with segmented images in the simulation environment. Next, we employ two U-Nets to segment robotic hand-view and side-view images during real robot control. These U-Net are pre-trained with synthetic RGB and segmentation masks in the simulation environment and fine-tuned with only 20 real images. We evaluate the robustness of the proposed framework in both simulation and real environments. Seg-CURL is robust to the texture, lighting, shadow, and camera position gap. Finally, our algorithm is tested on a real Baxter robot with a dark hand-view in the cube lifting task with a success rate of 16/20 in zero-shot transfer.

show abstract

Contrastive Domain Adaptation

Cited by 4 publications

References 37 publications

DeepAstroUDA: semi-supervised universal domain adaptation for cross-survey galaxy morphology classification and anomaly detection

DeepAstroUDA: semi-supervised universal domain adaptation for cross-survey galaxy morphology classification and anomaly detection

DeepAstroUDA: Semi-Supervised Universal Domain Adaptation for Cross-Survey Galaxy Morphology Classification and Anomaly Detection

Seg-CURL: Segmented Contrastive Unsupervised Reinforcement Learning for Sim-to-Real in Visual Robotic Manipulation

Contact Info

Product

Resources

About