More Like this

1. Automatic detection of novelty galaxies in digital sky survey data

   Margapuri, V. (March 2021, International journal of computer application)

   null (Ed.)

   Galaxy images of the order of multi-PB are collected as part of modern digital sky surveys using robotic telescopes. While there is a plethora of imaging data available, the majority of the images that are captured resemble galaxies that are “regular”, i.e., galaxy types that are already known and probed. However, “novelty" galaxy types, i.e., little-known galaxy types are encountered on occasion. The astronomy community shows paramount interest in the novelty galaxy types since they contain the potential for scientific discovery. However, since these galaxies are rare, the identification of such novelty galaxies is not trivial and requires automation techniques. Since these novelty galaxies are by definition, not known, supervised machine learning models cannot be trained to detect them. In this paper, an unsupervised machine learning method for automatic detection of novelty galaxies in large databases is proposed. The method uses a large set of image features weighted by their entropy. To handle the impact of self-similar novelty galaxies, the most similar galaxies are ranked-ordered. In addition, Bag of Visual Words (BOVW) is assimilated to the problem of detecting novelty galaxies. Each image in the dataset is represented as a set of features made up of key-points and descriptors. A histogram of the features is constructed and is leveraged to identify the neighbors of each of the images. Experimental results using data from the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS) show that the performance of the methods in detecting novelty galaxies is superior to other shallow learning methods such as one-class SVM, Local Outlier Factor, and K-Means, and also newer deep learning-based methods such as auto-encoders. The dataset used to evaluate the method is publicly available and can be used as a benchmark to test future algorithms for automatic detection of peculiar galaxies. 

   more » « less
2. MedShift: Automated Identification of Shift Data for Medical Image Dataset Curation

   https://doi.org/10.1109/JBHI.2023.3275104  

   Guo, Xiaoyuan; Gichoya, Judy Wawira; Trivedi, Hari; Purkayastha, Saptarshi; Banerjee, Imon (August 2023, IEEE Journal of Biomedical and Health Informatics)

   Automated curation of noisy external data in the medical domain has long been in high demand, as AI technologies need to be validated using various sources with clean, annotated data. Identifying the variance between internal and external sources is a fundamental step in curating a high-quality dataset, as the data distributions from different sources can vary significantly and subsequently affect the performance of AI models. The primary challenges for detecting data shifts are - (1) accessing private data across healthcare institutions for manual detection and (2) the lack of automated approaches to learn efficient shift-data representation without training samples. To overcome these problems, we propose an automated pipeline called MedShift to detect top-level shift samples and evaluate the significance of shift data without sharing data between internal and external organizations. MedShift employs unsupervised anomaly detectors to learn the internal distribution and identify samples showing significant shiftness for external datasets, and then compares their performance. To quantify the effects of detected shift data, we train a multi-class classifier that learns internal domain knowledge and evaluates the classification performance for each class in external domains after dropping the shift data. We also propose a data quality metric to quantify the dissimilarity between internal and external datasets. We verify the efficacy of MedShift using musculoskeletal radiographs (MURA) and chest X-ray datasets from multiple external sources. Our experiments show that our proposed shift data detection pipeline can be beneficial for medical centers to curate high-quality datasets more efficiently. 

   more » « less
3. Composite Active Learning: Towards Multi-Domain Active Learning with Theoretical Guarantees

   https://doi.org/10.1609/AAAI.V38I11.29119  

   Hao, Guang-Yuan; Huang, Hengguan; Wang, Haotian; Gao, Jie; Wang, Hao (March 2024, Proceedings of the AAAI Conference on Artificial Intelligence)

   Active learning (AL) aims to improve model performance within a fixed labeling budget by choosing the most informative data points to label. Existing AL focuses on the single-domain setting, where all data come from the same domain (e.g., the same dataset). However, many real-world tasks often involve multiple domains. For example, in visual recognition, it is often desirable to train an image classifier that works across different environments (e.g., different backgrounds), where images from each environment constitute one domain. Such a multi-domain AL setting is challenging for prior methods because they (1) ignore the similarity among different domains when assigning labeling budget and (2) fail to handle distribution shift of data across different domains. In this paper, we propose the first general method, dubbed composite active learning (CAL), for multi-domain AL. Our approach explicitly considers the domain-level and instance-level information in the problem; CAL first assigns domain-level budgets according to domain-level importance, which is estimated by optimizing an upper error bound that we develop; with the domain-level budgets, CAL then leverages a certain instance-level query strategy to select samples to label from each domain. Our theoretical analysis shows that our method achieves a better error bound compared to current AL methods. Our empirical results demonstrate that our approach significantly outperforms the state-of-the-art AL methods on both synthetic and real-world multi-domain datasets. Code is available at https://github.com/Wang-ML-Lab/multi-domain-active-learning. 

   more » « less
4. Constellation dataset: benchmarking high-altitude object detection for an urban intersection

   Turkcan, M; Narasimhan, S; Zang, C; Je, G; Yu, B; Ghasemi, M; Ghaderi, J; Zussman, G; Kostic, Z (April 2024, arXiv:2404.16944 [cs.CV], Apr. 2024.)

   We introduce Constellation, a dataset of 13K images suitable for research on detection of objects in dense urban streetscapes observed from high-elevation cameras, collected for a variety of temporal conditions. The dataset addresses the need for curated data to explore problems in small object detection exemplified by the limited pixel footprint of pedestrians observed tens of meters from above. It enables the testing of object detection models for variations in lighting, building shadows, weather, and scene dynamics. Weevaluate contemporary object detection architectures on the dataset, observing that state-of-the-art methods have lower performance in detecting small pedestrians compared to vehicles, corresponding to a 10% difference in average precision (AP). Using structurally similar datasets for pretraining the models results in an increase of 1.8% mean AP (mAP). We further find that incorporating domain-specific data augmentations helps improve model performance. Using pseudo-labeled data, obtained from inference outcomes of the best-performing models, improves the performance of the models. Finally, comparing the models trained using the data collected in two different time intervals, we find a performance drift in models due to the changes in intersection conditions over time. The best-performing model achieves a pedestrian AP of 92.0% with 11.5 ms inference time on NVIDIA A100 GPUs, and an mAP of 95.4%. 

   more » « less
5. Single-Shot Domain Adaptation via Target-Aware Generative Augmentations

   https://doi.org/10.1109/ICASSP49357.2023.10096784  

   Subramanyam, Rakshith; Thopalli, Kowshik; Berman, Spring; Turaga, Pavan; Thiagarajan, Jayaraman J. (June 2023, ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP))

   The problem of adapting models from a source domain using data from any target domain of interest has gained prominence, thanks to the brittle generalization in deep neural networks. While several test-time adaptation techniques have emerged, they typically rely on synthetic data augmentations in cases of limited target data availability. In this paper, we consider the challenging setting of single-shot adaptation and explore the design of augmentation strategies. We argue that augmentations utilized by existing methods are insufficient to handle large distribution shifts, and hence propose a new approach SiSTA (Single-Shot Target Augmentations), which first fine-tunes a generative model from the source domain using a single-shot target, and then employs novel sampling strategies for curating synthetic target data. Using experiments with a state-of-the-art domain adaptation method, we find that SiSTA produces improvements as high as 20% over existing baselines under challenging shifts in face attribute detection, and that it performs competitively to oracle models obtained by training on a larger target dataset. Our codes can be accessed at github.com/kowshikthopalli/SISTA. 

   more » « less