mlsanityData to 14 Oct 2026

Test-time augmentation-based active learning and self-training for label-efficient segmentation

arXiv 2308.10727 ↗First posted 21 Aug 2023cs.CV cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Deep learning techniques depend on large datasets whose annotation is time-consuming. To reduce annotation burden, the self-training (ST) and active-learning (AL) methods have been developed as well as methods that combine them in an iterative fashion. However, it remains unclear when each method is the most useful, and when it is advantageous to combine them. In this paper, we propose a new method that combines ST with AL using Test-Time Augmentations (TTA). First, TTA is performed on an initial teacher network. Then, cases for annotation are selected based on the lowest estimated Dice score. Cases with high estimated scores are used as soft pseudo-labels for ST. The selected annotated cases are trained with existing annotated cases and ST cases with border slices annotations. We demonstrate the method on MRI fetal body and placenta segmentation tasks with different data variability characteristics. Our results indicate that ST is highly effective for both tasks, boosting performance for in-distribution (ID) and out-of-distribution (OOD) data. However, while self-training improved the performance of single-sequence fetal body segmentation when combined with AL, it slightly deteriorated performance of multi-sequence placenta segmentation on ID data. AL was helpful for the high variability placenta data, but did not improve upon random selection for the single-sequence body data. For fetal body segmentation sequence transfer, combining AL with ST following ST iteration yielded a Dice of 0.961 with only 6 original scans and 2 new sequence scans. Results using only 15 high-variability placenta cases were similar to those using 50 cases. Code is available at: https://github.com/Bella31/TTA-quality-estimation-ST-AL

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 17 of its 27 references that are in the map. First posted on arXiv 21 Aug 2023.

Builds on

17 of 27 references.

References · 17

  1. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  2. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 510
  3. 2021Semi-Supervised Semantic Segmentation with Cross Pseudo Supervisioncited by 463
  4. 2020Semi-Supervised Semantic Segmentation with Cross-Consistency Trainingcited by 376
  5. 2020Rethinking Pre-training and Self-trainingcited by 352
  6. 2018Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networkscited by 280
  7. 2019Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentationcited by 260
  8. 2021ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentationcited by 184
  9. 2018Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysiscited by 179
  10. 2019A Survey on Active Learning and Human-in-the-Loop Deep Learning for Medical Image Analysiscited by 151
  11. 2017Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentationcited by 146
  12. 2020Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptationcited by 81
5 more
  1. 2019Assessing Reliability and Challenges of Uncertainty Estimations for Medical Image Segmentationcited by 75
  2. 2022BoostMIS: Boosting Medical Image Semi-supervised Learning with Adaptive Pseudo Labeling and Informative Active Annotationcited by 54
  3. 2022A Survey on Label-efficient Deep Image Segmentation: Bridging the Gap between Weak Supervision and Dense Predictioncited by 46
  4. 2022Warm Start Active Learning with Proxy Labels & Selection via Semi-Supervised Fine-Tuningcited by 21
  5. 2022Partial annotations for the segmentation of large structures with low annotation costcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2025SingleStrip: learning skull-stripping from a single labeled examplecited by 1
  2. 2026From Cold Start to Active Learning: Embedding-Based Scan Selection for Medical Image Segmentationcited by 1
  3. 2026Nested Radially Monotone Polar Occupancy Estimation: Clinically-Grounded Optic Disc and Cup Segmentation for Glaucoma Screeningcited by 0