Semantic Segmentation with Active Semi-Supervised Learning

Rangnekar, Aneesh; Kanan, Christopher; Hoffman, Matthew J.

doi:10.1109/wacv56688.2023.00591

Cited by 20 publications

(6 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The sample selection in AL is done using uncertainty (Liu et al 2019), entropy (Aghdam et al 2019), core-set selection (Sener and Savarese 2017) or mutual-information (Kirsch, Van Amersfoort, and Gal 2019). While there have been some prior works that combine AL and SSL for object detection and segmentation task (Elezi et al 2022;Rangnekar, Kanan, and Hoffman 2023), we are the first to propose a unified SSL active learning framework for spatio-temporal video action detection to best of our knowledge. We use data perturbation via noise based augmentation to get the model's uncertainty, using that as an estimate of usefulness for each sample in our AL strategy.…”

Section: Related Workmentioning

confidence: 99%

Semi-supervised Active Learning for Video Action Detection

Singh,

Rana,

Kumar

et al. 2024

AAAI

View full text Add to dashboard Cite

In this work, we focus on label efficient learning for video action detection. We develop a novel semi-supervised active learning approach which utilizes both labeled as well as un- labeled data along with informative sample selection for ac- tion detection. Video action detection requires spatio-temporal localization along with classification, which poses several challenges for both active learning (informative sample se- lection) as well as semi-supervised learning (pseudo label generation). First, we propose NoiseAug, a simple augmenta- tion strategy which effectively selects informative samples for video action detection. Next, we propose fft-attention, a novel technique based on high-pass filtering which enables effective utilization of pseudo label for SSL in video action detection by emphasizing on relevant activity region within a video. We evaluate the proposed approach on three different bench- mark datasets, UCF-101-24, JHMDB-21, and Youtube-VOS. First, we demonstrate its effectiveness on video action detec- tion where the proposed approach outperforms prior works in semi-supervised and weakly-supervised learning along with several baseline approaches in both UCF101-24 and JHMDB- 21. Next, we also show its effectiveness on Youtube-VOS for video object segmentation demonstrating its generalization capability for other dense prediction tasks in videos.

show abstract

Section: Related Workmentioning

confidence: 99%

Semi-supervised Active Learning for Video Action Detection

Singh,

Rana,

Kumar

et al. 2024

AAAI

View full text Add to dashboard Cite

show abstract

“…Many methods, such as EquAL [15], Ensemble+AT [24], and CEAL [36], estimate uncertainty based on the output probabilities. Epistemic uncertainty, estimated using Entropy [32], is often used a as strong baseline in several active learning works [15,29,33]. Some methods, namely BALD [17] and DBAL [12] employed a Bayesian approach using Monte Carlo Dropout [11] to measure the epistemic uncertainty.…”

Section: Deep Active Learningmentioning

confidence: 99%

Best Practices in Active Learning for Semantic Segmentation

Mittal,

Niemeijer,

Schäfer

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Active learning is particularly of interest for semantic segmentation, where annotations are costly. Previous academic studies focused on datasets that are already very diverse and where the model is trained in a supervised manner with a large annotation budget. In contrast, data collected in many driving scenarios is highly redundant, and most medical applications are subject to very constrained annotation budgets. This work investigates the various types of existing active learning methods for semantic segmentation under diverse conditions across three dimensions -data distribution w.r.t. different redundancy levels, integration of semi-supervised learning, and different labeling budgets. We find that these three underlying factors are decisive for the selection of the best active learning approach. As an outcome of our study, we provide a comprehensive usage guide to obtain the best performance for each case. It is the first systematic study that investigates these dimensions covering a wide range of settings including more than 3K model training runs. In this work, we also propose an exemplary evaluation task for driving scenarios, where data has high redundancy, to showcase the practical implications of our research findings.

show abstract

“…The highest-ranked samples, deemed the most informative ones, are sent to the human annotators for the AL step; the remaining pseudo-labels are kept for the next training. The mean teacher approach of [42] uses SSL in the teacher network to generate pseudo labels used to train the student network; the samples with the best performance are labeled and inserted in the labeled pool for the subsequent iteration, and the weights of the teacher network are updated as moving average of the weights of the student network.…”

Section: Combining Semi-supervised and Active Learningmentioning

confidence: 99%

Computational Evaluation of the Combination of Semi-Supervised and Active Learning for Histopathology Image Segmentation with Missing Annotations

Jiménez,

Dierckx,

Amodei

et al. 2023

2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)

View full text Add to dashboard Cite

Real-world segmentation tasks in digital pathology require a great effort from human experts to accurately annotate a sufficiently high number of images. Hence, there is a huge interest in methods that can make use of nonannotated samples, to alleviate the burden on the annotators. In this work, we evaluate two classes of such methods, semi-supervised and active learning, and their combination on a version of the GlaS dataset for gland segmentation in colorectal cancer tissue with missing annotations. Our results show that semi-supervised learning benefits from the combination with active learning and outperforms fully supervised learning on a dataset with missing annotations. However, an active learning procedure alone with a simple selection strategy obtains results of comparable quality.

show abstract

Semantic Segmentation with Active Semi-Supervised Learning

Cited by 20 publications

References 60 publications

Semi-supervised Active Learning for Video Action Detection

Semi-supervised Active Learning for Video Action Detection

Best Practices in Active Learning for Semantic Segmentation

Computational Evaluation of the Combination of Semi-Supervised and Active Learning for Histopathology Image Segmentation with Missing Annotations

Contact Info

Product

Resources

About