Pith. sign in

REVIEW 6 cited by

Unbiased Teacher for Semi-Supervised Object Detection

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2102.09480 v1 pith:BRVCVHEA submitted 2021-02-18 cs.CV cs.LG

Yen-Cheng Liu , Chih-Yao Ma , Zijian He , Chia-Wen Kuo , Kan Chen , Peizhao Zhang , Bichen Wu , Zsolt Kira
show 1 more author
Peter Vajda
This is my paper · ORCID
classification cs.CVcs.LG
keywords teacherunbiaseddatadetectionlabeledobjectsemi-supervisedachieves
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Semi-supervised learning, i.e., training networks with both labeled and unlabeled data, has made significant progress recently. However, existing works have primarily focused on image classification tasks and neglected object detection which requires more annotation effort. In this work, we revisit the Semi-Supervised Object Detection (SS-OD) and identify the pseudo-labeling bias issue in SS-OD. To address this, we introduce Unbiased Teacher, a simple yet effective approach that jointly trains a student and a gradually progressing teacher in a mutually-beneficial manner. Together with a class-balance loss to downweight overly confident pseudo-labels, Unbiased Teacher consistently improved state-of-the-art methods by significant margins on COCO-standard, COCO-additional, and VOC datasets. Specifically, Unbiased Teacher achieves 6.8 absolute mAP improvements against state-of-the-art method when using 1% of labeled data on MS-COCO, achieves around 10 mAP improvements against the supervised baseline when using only 0.5, 1, 2% of labeled data on MS-COCO.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SS-DC: Spatial-Spectral Decoupling and Coupling Across Visible-Infrared Gap for Domain Adaptive Object Detection

    cs.CV 2025-07 conditional novelty 6.0 of 10

    SS-DC improves RGB-to-infrared domain-adaptive object detection by spectrally decoupling domain-invariant from domain-specific features and coupling them with spatial features.

  2. De-Simplifying Pseudo Labels to Enhancing Domain Adaptive Object Detection

    cs.CV 2025-07 conditional novelty 6.0 of 10

    DeSimPL reduces the share of easy pseudo-labels during self-labeling domain-adaptive detection, improving SimROD by 2 to 5 mAP on four benchmarks.

  3. Dual Guidance Semi-Supervised Action Detection

    cs.CV 2025-07 conditional novelty 5.0 of 10

    Dual guidance, combining frame-level action classification with box-level prediction, improves pseudo-box selection for semi-supervised spatio-temporal action localization.

  4. Influence of color correction on pathology detection in Capsule Endoscopy

    cs.CV 2025-01 conditional novelty 5.0 of 10

    Color correction of capsule endoscopy images changes detection bounding boxes and false positive counts, but does not consistently improve standard detection metrics.

  5. PyKirigami: An interactive Python simulator for kirigami structures

    cond-mat.soft 2025-08 unverdicted novelty 4.0 of 10

    PyKirigami is an open-source Python simulator that models kirigami tessellations as articulated rigid-body networks for real-time deployment simulation, collision detection, and identification of geometric locking states.

  6. SST: Self-training with Self-adaptive Thresholding for Semi-supervised Learning

    cs.LG 2025-05 conditional novelty 4.0 of 10

    SST uses per-class thresholds updated once per cycle to pick pseudo-labels, reporting 84.9% ImageNet top-1 accuracy with 10% labeled data on a huge ViT.

Pith tools