Pith. sign in

REVIEW 3 major objections 5 minor 34 references

Visual Content Detection in Educational Videos with Transfer Learning and Dataset Enrichment

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Fine-tuning YOLOv11 on a dataset that combines 1,000 manually labeled lecture frames with 3,000 auto-labeled frames raises AP50 from 90.75% to 95.32% on the LVVO_1k benchmark, and the paper shows this semi-supervised enrichment makes a…

desk verdict Useful dataset and detector comparison, but the auto-labeling gain is likely inflated by test-fold leakage. read the letter →

arxiv 2506.21903 v2 pith:4EYXK6PH submitted 2025-06-27 cs.CV

Dipayan Biswas , Shishir Shah , Jaspal Subhlok This is my paper · ORCID
classification cs.CV
keywords educationalvideoanalysislectureframesvisualobjectdetectiontransferlearningsemi-supervisedauto-labelingdatasetenrichmentYOLOv11LVVObenchmark
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether a detector for visual elements in lecture slides—charts, tables, graphs, and illustrations—can be built with very few manual labels. The authors evaluate six off-the-shelf object detectors fine-tuned on three lecture-video datasets and find that YOLOv11 is the strongest. They then grow a 1,000-frame manually labeled corpus (LVVO_1k) into a 4,000-frame corpus by having a YOLO teacher label 3,000 additional frames, and show that progressively fine-tuning the model on the combined data raises AP50 from 90.75% to 95.32% and average precision from 77.6% to 84.19% on the manually labeled benchmark. The paper also shows that jointly training on all three datasets yields a single model nearly as accurate as dataset-specific models. A new publicly released benchmark, LVVO, accompanies the claims.

What carries the argument

The carrying mechanism is the combination of transfer learning and a semi-supervised auto-labeling pipeline. A COCO-pretrained YOLOv11 (a single-stage real-time object detector) is first fine-tuned on the manually annotated LVVO_1k; this teacher model then generates bounding boxes for the 3,000 unlabeled frames with a 0.5 confidence filter, producing LVVO_3k. The student is trained on the union by progressive fine-tuning, starting from the teacher's weights rather than from COCO weights, which the paper finds superior to training from COCO on the combined set. The evaluation is 5-fold cross-validation on LVVO_1k using COCO metrics AP, AP50, and AP75.

What would settle it

Run the auto-labeling pipeline inside each cross-validation fold, training the teacher only on that fold's training images and labeling only that fold's validation images, then evaluate the student on those same validation images; if the AP50 gain shrinks substantially from the reported 6.59-point increase, the improvement is an artifact of the teacher having seen the test fold.

Watch

Extended reading notes

Core claim

The central claim is that the bottleneck for detecting visual objects in lecture videos is not the detector architecture but the scarcity of labeled frames, and that this scarcity can be mitigated by a semi-supervised auto-labeling loop. Starting from a COCO-pretrained YOLOv11, the authors fine-tune on 1,000 manually annotated LVVO frames using an 80/20 split, use that model to label 3,000 unlabeled frames at a confidence threshold of 0.5, and then continue fine-tuning the same model on the full 4,000-image set (progressive fine-tuning). On the 1,000 manually annotated frames, this raises AP50 from 90.75% to 95.32% and AP from 77.6% to 84.19%. The paper further claims that a model trained jointly on the LDD, LPM, and LVVO_1k datasets performs comparably to models trained on a single dataset, suggesting a general detector is achievable with a modest diverse sample.

Load-bearing premise

The method assumes that the auto-generated bounding boxes for the 3,000 unlabeled frames are accurate enough to teach the detector, and that measuring the gain on the same 1,000 manually labeled frames that helped create those pseudo-labels is a fair test of that gain.

Editorial extensions

If this is right

  • Lecture-video search, navigation, and summarization tools can obtain usable detectors with 1,000 manual labels plus a few thousand unlabeled frames, avoiding large annotation budgets.
  • A single model trained jointly on LDD, LPM, and LVVO matches dataset-specific models, so a general-purpose lecture-video detector is within reach from a few hundred images per domain.
  • The biggest gain from auto-labeled data comes at the first 1,000 added frames, so enrichment campaigns can stop early without losing most of the benefit.
  • Fine-tuning on related in-domain datasets helps mainly when the target labeled set is small (below roughly 60% of 1,000 frames), giving a concrete recipe for low-resource lecture-video projects.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the auto-labeling result holds under nested cross-validation, the same teacher-labeler loop should transfer to other structured document images, such as scanned worksheets or whiteboard photos, where objects are also artificial and boundary-free.
  • The confidence threshold of 0.5 for pseudo-labels is fixed without a sensitivity study; sweeping this threshold would reveal the precision-recall trade-off of the synthesized labels and likely an optimal operating point.
  • The diminishing-returns curve for added auto-labeled data suggests an adaptive enrichment budget: stop adding pseudo-labeled frames when validation AP stops improving.
  • A direct comparison of progressive fine-tuning against training the student from COCO weights on the same combined data would isolate how much of the reported gain comes from the teacher's initialization rather than from the extra data alone.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper addresses detection of visual objects (charts, tables, illustrations) in educational lecture videos. It introduces the LVVO dataset (4,000 frames, of which 1,000 are manually annotated), evaluates six COCO-pretrained object detectors on lecture video datasets, identifies YOLOv11 as the best performer, explores cross-dataset and joint-training strategies, and proposes a semi-supervised auto-labeling pipeline that enriches the training set with 3,000 automatically labeled LVVO frames. The authors report that progressive fine-tuning with the enriched dataset raises AP50 from 90.75% to 95.32% and AP from 77.6% to 84.19% on the LVVO_1k benchmark.

Significance. If the enrichment result were independent, the paper would offer a valuable practical recipe for leveraging unlabeled lecture video frames and a new benchmark (LVVO) for the community. The systematic comparison of multiple detectors on three datasets and the release of the dataset and code are concrete contributions. However, the central claim of dataset enrichment is compromised by an evaluation-leakage problem in Section V-C: the auto-labeling teacher is trained on 80% of LVVO_1k and the enriched models are then evaluated by 5-fold cross-validation on the same LVVO_1k, so the teacher has already seen the test-fold images and their annotation style. The reported gains therefore reflect consistency with the teacher's training data rather than genuine generalization to new lecture content. The cross-dataset and joint-training experiments are less affected and retain value, but the headline enrichment claim needs to be re-established under a leakage-free protocol.

major comments (3)
  1. [Section V-C, Table II] The dataset enrichment evaluation has a structural leakage problem. The auto-labeling teacher is fine-tuned on an 80% split of LVVO_1k (Section V-C) and then used to generate pseudo-labels for LVVO_3k. The resulting enriched models are evaluated using 5-fold cross-validation on the same LVVO_1k set. Because the teacher's 800-image training set overlaps with four of the five 200-image test folds (expected overlap roughly 160 images per fold), the teacher's weights encode the appearance and annotation style of the validation folds. Training the Progressive or Comprehensive model on pseudo-labels produced by this teacher and then measuring AP50 on those same folds does not provide an independent estimate of the enrichment benefit. The 6.59-point AP50 gain reported in Table II is therefore not trustworthy as a measure of generalization to new lecture video content. The authors should redo the enrichment experiment with a held-out split that is disjoint from the teacher's training set, or with a teacher trained only on the per-fold training splits when generating pseudo-labels for enrichment, and report the resulting performance.
  2. [Section V-C, Figure 7] The pseudo-label quality is not assessed at all. The manuscript applies a confidence threshold of 0.5 to filter auto-labeled detections but provides no statistics on the resulting LVVO_3k labels: no distribution of detections per image, no agreement with manual labels on a subset, and no sensitivity analysis of the threshold. Since the whole enrichment claim rests on the assumption that these pseudo-labels are accurate enough to serve as training annotations, the paper should at least report a small-scale manual evaluation of the auto-labeled boxes or a comparison of models trained with different confidence thresholds. Without such evidence, the reader cannot separate the effect of added data from the effect of label noise.
  3. [Section IV versus Section V] The evaluation protocols are inconsistent across the paper's main claims. Section IV evaluates all models on a single 80%-20% train-validation split, while Section V (Figures 3 and 5, Table II) uses 5-fold cross-validation. The numbers in Figure 2 (e.g., YOLOv11 AP50 on LVVO_1k) are not directly comparable to the Baseline row in Table II, which is obtained with 5-fold CV. This makes it difficult to track how the model's performance evolves from the initial comparison to the enriched setting. Please state explicitly which protocol is used for each result and, where possible, report both metrics under the same protocol.
minor comments (5)
  1. [Section II] There is a typo in the sentence "We focus on locating visual elements not clasification." It should read "We focus on locating visual elements, not classification."
  2. [Throughout] The model name is written inconsistently as "YOLO11" (e.g., Figure 3 caption) and "YOLOv11" elsewhere. Please standardize to "YOLOv11" throughout.
  3. [Section V] The terms "Fine-tuning" and "Finetuning" appear with inconsistent capitalization and hyphenation (e.g., "Progressive Finetuning" vs. "Progressive Fine-Tuning"). Please unify the terminology.
  4. [Figure 5] The y-axis labels in Figure 5 are missing; the reader cannot tell whether the bars represent AP50 or overall AP for the left and right panels. Please add axis labels and a legend explaining "yellow" and "red" bars in the caption.
  5. [Section III] The paper states that LVVO has four visual categories but also says category labels are not used. Please clarify whether the four categories are used for the detection task or only for the annotation process, and whether the reported metrics are class-agnostic or class-specific.

Circularity Check

1 steps flagged · score 6.0 of 10

Enrichment gain in Table II is fitted-input-called-prediction: the auto-labeling teacher is trained on 80% of LVVO_1k, so its pseudo-labels leak the 5-fold CV test folds of the same LVVO_1k images.

  1. fitted input called prediction [Section V-C, Table II]
    "Specifically, we fine-tune a COCO-pretrained YOLOv11 model using transfer learning, first adapting it to the manually annotated LVVO_1k dataset, which was divided into 80% for training and 20% for validation. ... Standard 5-fold cross-validation was applied for this baseline and for the following two fine-tuning approaches ... Table II summarizes the performance results on the same cross-validation images of manually labeled LVVO 1k dataset."

    The auto-labeling teacher is fine-tuned on an 80% subset of LVVO_1k; then the enriched model is trained on the resulting LVVO_3k pseudo-labels and evaluated by 5-fold CV on the same LVVO_1k images. Since the teacher's 800 training images and the five 200-image folds are subsets of the same 1,000 images, each held-out fold overlaps the teacher's training set by 160 images on average. The pseudo-labels therefore encode the teacher's weights and annotation style on the very images used for evaluation. The Table II gain (90.75 to 95.32 AP50) measures the student's consistency with a teacher that has already seen most of the evaluation images, not the value of new information in the unlabeled LVVO_3k frames.

full rationale

The paper's other empirical contributions are self-contained and not circular: the model comparison on three datasets, the cross-dataset transfer experiments, and the joint-training generalization study compare against external benchmarks and do not depend on the auto-labeling loop. However, the headline dataset-enrichment claim (contribution 3, Table II) is compromised by a closed loop between the teacher model and the evaluation set. The teacher that generates LVVO_3k pseudo-labels is fine-tuned on 80% of LVVO_1k, and the benefit of those pseudo-labels is then measured by 5-fold cross-validation on the same LVVO_1k images. Since the teacher's training images and the CV folds partition the same 1,000 frames, the pseudo-labels carry the teacher's already-seen test-frame information into the student's training. The reported AP50 improvement from 90.75% to 95.32% is therefore not an independent measure of generalization to new lecture content; it is, in part, a teacher-student consistency check on a distribution the teacher was fitted to. This is a partial circularity, not a complete one, because the pseudo-labels themselves come from different images and the general-model and cross-dataset results remain independent. Score 6 reflects that one central claim reduces by construction to the teacher's prior exposure to the evaluation set.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim depends primarily on standard transfer-learning hyperparameters and on two domain assumptions: that bounding-box representation is adequate for lecture video visual objects, and that auto-labels are trustworthy. The most consequential assumption is the accuracy of the pseudo-labels, which is neither manually validated nor independently tested. No new entities are postulated.

free parameters (4)
  • auto-labeling confidence threshold = 0.5
    Used in Section V-C to filter low-confidence predictions for pseudo-labels; directly determines the LVVO_3k training set and influences the reported accuracy gain.
  • learning rate = 0.001
    Global learning rate for all fine-tuning experiments, chosen empirically (Section IV).
  • batch size = 8
    Batch size for all experiments, chosen empirically (Section IV).
  • training epochs = 30
    Number of epochs for all fine-tuning, chosen empirically (Section IV).
assumptions (3)
  • domain assumption Axis-aligned bounding boxes and COCO IoU metrics adequately represent visual objects in lecture videos.
    The paper assumes visual elements can be localized with rectangular boxes and evaluated by IoU, despite stating in Section I that object boundaries are context-dependent and lack clear physical separation.
  • ad hoc to paper Pseudo-labels generated by a YOLO model at 0.5 confidence are accurate enough to serve as training annotations.
    Central to dataset enrichment in Section V-C; no manual verification of the 3,000 auto-labeled frames is performed.
  • domain assumption Frames selected based on visual richness are representative of lecture videos in general.
    Section III describes selecting frames based on visual richness and filtering near-duplicates, which may bias the dataset toward visually busy frames and limit generalizability to other lecture video collections.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Visual Content Detection in Educational Videos with Transfer Learning and Dataset Enrichment." pith.science (2026). https://pith.science/paper/4EYXK6PH

@misc{pith2026250621903,
  author       = {Pith},
  title        = {Pith review of: Visual Content Detection in Educational Videos with Transfer Learning and Dataset Enrichment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4EYXK6PH}},
  note         = {Machine review of arXiv:2506.21903}
}
read the original abstract

Video is transforming education with online courses and recorded lectures supplementing and replacing classroom teaching. Recent research has focused on enhancing information retrieval for video lectures with advanced navigation, searchability, summarization, as well as question answering chatbots. Visual elements like tables, charts, and illustrations are central to comprehension, retention, and data presentation in lecture videos, yet their full potential for improving access to video content remains underutilized. A major factor is that accurate automatic detection of visual elements in a lecture video is challenging; reasons include i) most visual elements, such as charts, graphs, tables, and illustrations, are artificially created and lack any standard structure, and ii) coherent visual objects may lack clear boundaries and may be composed of connected text and visual components. Despite advancements in deep learning based object detection, current models do not yield satisfactory performance due to the unique nature of visual content in lectures and scarcity of annotated datasets. This paper reports on a transfer learning approach for detecting visual elements in lecture video frames. A suite of state of the art object detection models were evaluated for their performance on lecture video datasets. YOLO emerged as the most promising model for this task. Subsequently YOLO was optimized for lecture video object detection with training on multiple benchmark datasets and deploying a semi-supervised auto labeling strategy. Results evaluate the success of this approach, also in developing a general solution to the problem of object detection in lecture videos. Paper contributions include a publicly released benchmark of annotated lecture video frames, along with the source code to facilitate future research.

Figures

Figures reproduced from arXiv: 2506.21903 by the authors.

Figure 1
Figure 1. Sample lecture video frames with conceptual visual objects bounded by dotted green rectangles [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Average Precision (AP50%) comparison of different object detection [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Performance comparison between the Logiform algorithm [16] and YOLOv11 on the LVVO 1k dataset using 5-fold cross-validation [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Performance of object detection models trained on one dataset (labels [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 7
Figure 7. Figure 7: Different fine-tuning approaches using the extended dataset. Where, [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]
Figure 8
Figure 8. Figure 8: Performance improvement across AP50, AP75, and AP metrics as 1k, [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 28 canonical work pages

  1. [1]

    How video production affects student engagement: An empirical study of mooc videos,

    P. J. Guo, J. Kim, and R. Rubin, “How video production affects student engagement: An empirical study of mooc videos,” in Proceedings of the first ACM conference on Learning@ scale conference , 2014, pp. 41–50

    work page 2014

  2. [2]

    Students’ and instructors’ use of massive open online courses (moocs): Motivations and challenges,

    K. F. Hew and W. S. Cheung, “Students’ and instructors’ use of massive open online courses (moocs): Motivations and challenges,” Educational research review, vol. 12, pp. 45–58, 2014

    work page 2014

  3. [3]

    Student perceptions of indexed, searchable videos of faculty lectures,

    L. Barker, C. L. Hovey, J. Subhlok, and T. Tuna, “Student perceptions of indexed, searchable videos of faculty lectures,” in Proceedings of the 44th Annual Frontiers in Education Conference(FIE) , Madrid, Spain, Oct 2014

    work page 2014

  4. [4]

    Information retention from powerpoint™ and traditional lectures,

    A. Savoy, R. W. Proctor, and G. Salvendy, “Information retention from powerpoint™ and traditional lectures,” Computers & Education, vol. 52, no. 4, pp. 858–867, 2009

    work page 2009

  5. [5]

    Powerpoint’s power in the classroom: Enhancing students’ self-efficacy and attitudes,

    J. E. Susskind, “Powerpoint’s power in the classroom: Enhancing students’ self-efficacy and attitudes,” Computers & education , vol. 45, no. 2, pp. 203–215, 2005

    work page 2005

  6. [6]

    How the design of presentation slides affects audience comprehension: A case for the assertion-evidence approach,

    J. Garner and M. Alley, “How the design of presentation slides affects audience comprehension: A case for the assertion-evidence approach,” International Journal of Engineering Education , vol. 29, no. 6, pp. 1564–1579, 2013

    work page 2013

  7. [7]

    Aids to computer-based multimedia learning,

    R. E. Mayer and R. Moreno, “Aids to computer-based multimedia learning,” Learning and instruction , vol. 12, no. 1, pp. 107–119, 2002

    work page 2002

  8. [8]

    Animations need narrations: An experimental test of a dual-coding hypothesis

    R. E. Mayer and R. B. Anderson, “Animations need narrations: An experimental test of a dual-coding hypothesis.” Journal of educational psychology, vol. 83, no. 4, p. 484, 1991

    work page 1991

Show all 34 references
  1. [9]

    Cognitive principles of multimedia learning: The role of modality and contiguity

    R. Moreno and R. E. Mayer, “Cognitive principles of multimedia learning: The role of modality and contiguity.” Journal of educational psychology, vol. 91, no. 2, p. 358, 1999

    1999

  2. [10]

    Vizig: Anchor points based non-linear navigation and summarization in educational videos,

    K. Yadav, A. Gandhi, A. Biswas, K. Shrivastava, S. Srivastava, and O. Deshmukh, “Vizig: Anchor points based non-linear navigation and summarization in educational videos,” in Proceedings of the 21st Inter- national Conference on Intelligent User Interfaces , 2016, pp. 407–418

    2016

  3. [11]

    Enhancing lecture video navigation with AI generated summaries,

    M. R. Rahman, R. S. Koka, S. K. Shah, T. Solorio, and J. Subhlok, “Enhancing lecture video navigation with AI generated summaries,” Education and Information Technologies , pp. 1–24, 2023

    2023

  4. [12]

    Fastperson: Enhancing video-based learning through video summarization that preserves linguistic and visual contexts,

    K. Kawamura and J. Rekimoto, “Fastperson: Enhancing video-based learning through video summarization that preserves linguistic and visual contexts,” in Proceedings of the Augmented Humans International Conference 2024, 2024, pp. 205–216

    2024

  5. [13]

    Fitvid: Responsive and flexible video content adaptation,

    J. Kim, Y . Choi, M. Kahng, and J. Kim, “Fitvid: Responsive and flexible video content adaptation,” in Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems , 2022, pp. 1–16

    2022

  6. [14]

    Say it all: Feedback for improving non-visual presentation accessibility,

    Y .-H. Peng, J. Jang, J. P. Bigham, and A. Pavel, “Say it all: Feedback for improving non-visual presentation accessibility,” in Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems , 2021, pp. 1–12

    2021

  7. [15]

    Lecture2note: Automatic generation of lecture notes from slide-based educational videos,

    C. Xu, R. Wang, S. Lin, X. Luo, B. Zhao, L. Shao, and M. Hu, “Lecture2note: Automatic generation of lecture notes from slide-based educational videos,” in 2019 IEEE International Conference on Multi- media and Expo (ICME) . IEEE, 2019, pp. 898–903

    2019

  8. [16]

    Identification of visual objects in lecture videos with color and keypoints analysis,

    D. Biswas, S. Shah, and J. Subhlok, “Identification of visual objects in lecture videos with color and keypoints analysis,” in IEEE International Symposium on Multimedia (ISM) . IEEE, 2023, pp. 315–320

    2023

  9. [17]

    Semantic navigation of powerpoint-based lecture video for autonote generation,

    C. Xu, W. Jia, R. Wang, X. He, B. Zhao, and Y . Zhang, “Semantic navigation of powerpoint-based lecture video for autonote generation,” IEEE Transactions on Learning Technologies , vol. 16, no. 1, pp. 1–17, 2022

    2022

  10. [18]

    Enhancing lecture video navigation with AI generated summaries,

    M. R. Rahman, R. S. Koka, S. K. Shah, T. Solorio, and J. Subhlok, “Enhancing lecture video navigation with AI generated summaries,” Education and Information Technologies, vol. 29, no. 6, pp. 7361–7384,

  11. [19]

    Automatic explanation spot estimation method targeted at text and figures in lecture slides

    S. Tsujimura, K. Yamamoto, and S. Nakagawa, “Automatic explanation spot estimation method targeted at text and figures in lecture slides.” in INTERSPEECH, 2017, pp. 2764–2768

    2017

  12. [20]

    Videopoints: Lecture video platform,

    VideoPoints, “Videopoints: Lecture video platform,” https://videopoints. org, 2025, accessed: March 20, 2025

    2025

  13. [21]

    Lecture presentations multimodal dataset: Towards understanding multimodality in educational videos,

    D. W. Lee, C. Ahuja, P. P. Liang, S. Natu, and L.-P. Morency, “Lecture presentations multimodal dataset: Towards understanding multimodality in educational videos,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 20 087–20 098

    2023

  14. [22]

    Lecture video visual objects (lvvo) dataset: A benchmark for visual object detection in educational videos,

    D. Biswas, S. Shah, and J. Subhlok, “Lecture video visual objects (lvvo) dataset: A benchmark for visual object detection in educational videos,” arXiv preprint arXiv:2506.13657 , 2025

    2025 arXiv

  15. [23]

    Microsoft coco: Common objects in context,

    T.-Y . Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Doll ´ar, and C. L. Zitnick, “Microsoft coco: Common objects in context,” in Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13 . Springer,...

    2014

  16. [24]

    Faster r-cnn: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,” IEEE transactions on pattern analysis and machine intelligence, vol. 39, no. 6, pp. 1137–1149, 2016

    2016

  17. [25]

    Ssd: Single shot multibox detector,

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y . Fu, and A. C. Berg, “Ssd: Single shot multibox detector,” in Computer Vision– ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14 . Springer, 2016, pp. 21–37

    2016

  18. [26]

    Mask r-cnn,

    K. He, G. Gkioxari, P. Doll ´ar, and R. Girshick, “Mask r-cnn,” in Proceedings of the IEEE international conference on computer vision , 2017, pp. 2961–2969

    2017

  19. [27]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” in Proceedings of the IEEE international conference on computer vision , 2017, pp. 2980–2988

    2017

  20. [28]

    Fully convolutional one-stage 3d object detection on lidar range images,

    Z. Tian, X. Chu, X. Wang, X. Wei, and C. Shen, “Fully convolutional one-stage 3d object detection on lidar range images,” Advances in neural information processing systems , vol. 35, pp. 34 899–34 911, 2022

    2022

  21. [29]

    YOLOv11 Documentation,

    Ultralytics, “YOLOv11 Documentation,” 2024, accessed: 2 March

    2024

  22. [30]

    How transferable are features in deep neural networks?

    J. Yosinski, J. Clune, Y . Bengio, and H. Lipson, “How transferable are features in deep neural networks?” Advances in neural information processing systems, vol. 27, 2014

    2014

  23. [31]

    Meta-retinanet for few-shot object detection

    S. Li, W. Song, S. Li, A. Hao, and H. Qin, “Meta-retinanet for few-shot object detection.” in BMVC, 2020

    2020

  24. [32]

    Ultralytics yolo11,

    G. Jocher and J. Qiu, “Ultralytics yolo11,” 2024. [Online]. Available: https://github.com/ultralytics/ultralytics

    2024

  25. [2024]

    Available: https://doi.org/10.1007/s10639-023-11866-7

    [Online]. Available: https://doi.org/10.1007/s10639-023-11866-7

    doi

  26. [2025]

    Available: https://docs.ultralytics.com/models/yolo11

    [Online]. Available: https://docs.ultralytics.com/models/yolo11

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.