Pith. sign in

REVIEW 4 major objections 4 minor 39 references

DEEGITS: Deep Learning based Framework for Measuring Heterogenous Traffic State in Challenging Traffic Scenarios

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read A deep-learning framework fuses two public datasets to detect 14 road-user classes and measure traffic flow and speed in congested mixed traffic with the accuracy of manual counting.

desk verdict Useful applied ITS paper with a genuinely new fused dataset and honest field validation, but the speed equation as printed has wrong units and the calibration vectors don't match the declared tuple, so the speed pipeline is not reproducible without corrections. read the letter →

arxiv 2411.08335 v1 pith:M5546AFL submitted 2024-11-13 cs.CV cs.AI

classification cs.CVcs.AI
keywords trafficstatemeasurementheterogeneousmixedYOLOv8DeepSORTdatafusionvehicledetectionpedestrian
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

The paper tries to show that a single video-based deep-learning pipeline can measure traffic flow and speed in congested, heterogeneous road traffic, where motorized and non-motorized vehicles mix, with accuracy statistically indistinguishable from manual counting. To do this, it fuses two public datasets, one for pedestrians and one for local vehicles, into a new labelled dataset called DhakaPersons, and trains a YOLOv8 detector on 14 road-user classes. It then couples the detector with DeepSORT tracking and a geometric correction step to convert trajectories into flow and speed estimates. Field tests at two Dhaka intersections report correlations of 0.99–0.88 for flow and 0.91–0.97 for speed, with t-values below the 95% critical threshold. A sympathetic reader would take the central message to be that dataset fusion plus transfer learning is enough to make vision-based traffic-state measurement work in mixed traffic that standard benchmarks ignore.

What carries the argument

The load-bearing mechanism is the DhakaPersons fused dataset, built by cross-labelling CityPersons and DhakaAI with detectors trained on each, then manually correcting and merging the annotations into 14 classes based on the registered vehicle taxonomy. Around it sit a transfer-learned YOLOv8 anchor-free detector trained at 1280x1280 with SGD, the DeepSORT tracker (Kalman-filter motion model plus cosine appearance distance and Hungarian association), and a camera-skew correction that maps pixel trajectories into geodetic coordinates. These components feed two measurement operations: line-of-interest crossing counts for flow and per-track frame-to-frame distance accumulation for speed.

What would settle it

Take a fresh set of videos from a third Dhaka intersection with mixed motorized and non-motorized traffic, have two human annotators independently count vehicles and pedestrians, then run the DEEGITS pipeline; if the framework's per-minute flow and speed measurements differ from the human counts by more than the t-critical value of 2.015 at 95% confidence, the generality of the measurement claim fails.

Watch

Extended reading notes

Core claim

The central claim is that complementary data fusion solves the annotation gap that blocks simultaneous detection of pedestrians and vehicles in mixed traffic: a detector trained on DhakaAI labels vehicles in CityPersons images, a detector trained on CityPersons labels pedestrians in DhakaAI images, and manual correction merges the results into the 14-class DhakaPersons dataset (4,576 images, 59,576 boxes). On that dataset, a transfer-learned YOLOv8 model with SGD and grid-searched hyperparameters achieves 0.794 mAP@0.5 on validation and 0.786 mAP@0.5 on test, surpassing the prior DhakaAI benchmark of 0.458. When the detections are passed through DeepSORT and the trajectories are corrected for camera skew and crossed against a line of interest, the resulting flow and speed measurements at two urban signalized intersections, one motorized-dominant and one non-motorized-dominant, are statistically indistinguishable from manual ground truth under two-tailed t-tests at 95% confidence.

Load-bearing premise

The DhakaPersons ground-truth labels built by cross-labelling two datasets and manually correcting them are accurate enough to train and evaluate the detector, especially for rare classes like Ambulance, Special Purpose Vehicle, and Bicycle that have only 84 to 913 instances.

Editorial extensions

If this is right

  • Simultaneous detection of pedestrians plus 13 local vehicle classes is achievable on a fused dataset, and the resulting detector beats prior DhakaAI-specific benchmarks.
  • Traffic-flow and speed measurements from the framework are statistically indistinguishable from manual ground truth at 95% confidence at the two test sites, supporting automated alternative to manual counting.
  • The data-fusion recipe transfers to other regions because the class taxonomy is chosen from a local registry but the method does not depend on that particular taxonomy.
  • Framework accuracy degrades with severe occlusion and non-standard vehicles, naming the limiting conditions for practical deployment.
  • SGD generalizes better than Adam for this detection task, indicating that optimizer choice matters for avoiding overfitting on datasets with rare classes.

Reading between the lines

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

  • The aggregated t-test may conceal per-vehicle or per-class biases; a per-class or per-track error decomposition could reveal systematic undercounts, especially for rare classes, even when total flow looks unbiased.
  • The same cross-labelling fusion could be applied to other under-annotated domain pairs, such as night-time imagery or drone footage, where no existing dataset covers all target classes.
  • Because Ambulance and Special Purpose Vehicle have very few instances, class-weighted losses or synthetic augmentation could raise their reported 0.50 test accuracy and would be a direct test of the data-hunger hypothesis.
  • Tracking ID switches after occlusion, which the paper names as the cause of speed error at the first location, suggest that re-identification after occlusion would improve speed measurement more than further detector improvements.
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

4 major / 4 minor

Summary. DEEGITS is an end-to-end framework that combines a YOLOv8 detector fine-tuned on a new fused dataset (DhakaPersons), DeepSORT tracking, and camera-based geometric correction to measure traffic flow and speed at congested mixed-traffic intersections in Dhaka. The detector is trained on 14 BRTA-aligned vehicle classes plus pedestrians and reaches 0.794 mAP@0.5 on validation and 0.786 on test. Field tests at two signalized intersections are compared against manual ground truth using RMSE, Pearson correlation, and t-tests; the paper reports statistically insignificant differences.

Significance. The paper addresses a genuine gap by fusing vehicle-centric DhakaAI with pedestrian-centric CityPersons to enable simultaneous detection of heterogeneous traffic, and it validates the traffic-state pipeline against independent manual ground truth. Detection evaluation is standard and the validation-test gap is small. However, the central speed-measurement claim is not reproducible from the printed equations and calibration parameters, and the statistical evidence is incomplete; these issues must be resolved before the contribution can be fully credited.

major comments (4)
  1. [Traffic State Measurement, Eq. (12)] Equation (12) is dimensionally inconsistent as printed. If t_l are frame indices, then t_f - t_1 is a frame count and the denominator (t_f - t_1) * f has units of frames^2/second, so v(i)^k cannot be in m/s or km/h. With f = 25 fps, the printed denominator is 625 times larger than the correct time denominator (t_f - t_1)/f, which would make the reported speed RMSEs of 1.78 and 1.61 km/h impossible to obtain. If the implementation used a corrected denominator, the paper must state this explicitly; otherwise the speed validation is unsupported.
  2. [Traffic State Measurement, Eqs. (6)-(9); Field Validation] The camera-calibration description is not reproducible as printed. The parameter vector is defined as pi = {phi, omega, delta, X0, Y0}, but the reported pi1 has four values and pi2 has six values. In addition, Eq. (7) uses Y in the term phi * cot(delta) * Y, while the surrounding definitions indicate that the uncalibrated skewed coordinate y is intended, and phi and omega are defined as dimensionless magnification factors despite being used with geodetic coordinates. These inconsistencies prevent an independent implementation of the geometric correction that underpins the speed measurements.
  3. [Field Validation] The t-test evidence for statistical indistinguishability is incomplete. The paper reports t-values, p-values, and t-critical = 2.015, but it does not report the number of observations, degrees of freedom, or whether the test is paired or two-sample. This matters because the t-critical value depends on the degrees of freedom; for example, a paired two-tailed test with 12 five-minute intervals would have a critical value near 2.201, not 2.015. Without N and df, the claimed p-values cannot be verified.
  4. [Dataset Construction (DhakaPersons) and Table 1] The DhakaPersons test labels are partly generated by YOLOv8 models trained on DhakaAI and CityPersons, and although the authors state that mislabels were manually corrected, no audit protocol or inter-annotator check is reported. The rare classes show low diagonal accuracy in Table 1 (Ambulance 0.50, Special Purpose Vehicle 0.50 on the test set), so the overall mAP of 0.786 may overstate the reliability of the classification for the classes that matter most in mixed traffic. Adding an independently human-annotated hold-out sample, or at least per-class precision/recall on such a sample, would make the detection claim more robust.
minor comments (4)
  1. [Model Training / Hyperparameter Tuning] The paper refers to a subsection titled 'Hyperparameter Tuning' in the field-validation section, but no such subsection exists in the manuscript; the cross-reference should be added or removed.
  2. [Data Preparation and Analysis] The augmented-training-set arithmetic is not transparent: the paper states 3200 training images, states that each training example produces three outputs, and then reports 9674 images after augmentation including 50 background images, while 3200 * 3 + 50 = 9650. The discrepancy should be explained.
  3. [Traffic State Measurement, Eq. (12)] The notation in Eq. (12) should clarify whether t_l are frame indices or elapsed times, and whether F indexes all tracked frames or only the frames used in the interval; this is essential for reproducing the speed calculation.
  4. [Field Validation] The paper should state the time-interval length T_i and the number of intervals used in Figures 6 and 7, since the reported correlations and residuals otherwise cannot be interpreted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the detection mAP is scored against a held-out test split with manually corrected labels, and the traffic-state claims are validated against independent manual ground truth.

full rationale

Every load-bearing result in DEEGITS is checked against evidence that is not an input to the construction. The detection mAP is reported on a held-out 15% test split of the fused DhakaPersons dataset, and although that dataset is produced by mutual pseudo-labeling between YOLOv8 models trained on DhakaAI and CityPersons, the paper states that "The mislabeled annotations are manually corrected and similar type vehicles are merged into single classes," so the final test labels are not just the model's own outputs recycled by construction. The traffic-state measurements are validated against manual ground truth taken independently from the videos ("ground truth data (speed and flow) has also been collected from the video through manual post-processing"), and the reported RMSEs, Pearson correlations, t-values, and p-values are direct comparisons of the framework's measured quantities to that external ground truth, not fitted parameters disguised as predictions. Reference [39] is a self-citation, but it only supports the generic statement that flow and speed are the mostly used traffic-state measurement variables; it is not load-bearing. The paper's own limitations (severe occlusions, foreign vehicle types, underrepresented classes, and no nighttime validation) are explicit and do not create circularity. The printed unit ambiguity in Eq. (12) and the mismatched lengths of the reported calibration vectors are reproducibility and correctness defects, not instances of a result being equal to its input by construction.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claims rest on standard supervised learning plus a pseudo-labeled fused dataset and a small field validation. The trained weights and hyperparameters are fitted values, and the camera calibration and manual ground truth are measured inputs. No new physical entities are proposed.

free parameters (6)
  • YOLOv8 model weights (yolov8s.pt fine-tuned) = not released
    Learned from 9,674 augmented training images and pretrained COCO weights; the central detection accuracy depends on these weights.
  • Grid-search hyperparameters = lr=0.01, momentum=0.9, weight decay=0.0005, batch=12, epochs=100, imgsz=1280
    Chosen by Grid Search on validation; SGD selected over Adam and RMSProp because it reduced overfitting.
  • Data fusion labeling thresholds = minimum 5 pedestrian annotations per image; class-agnostic NMS; manual correction
    Hand-chosen criteria for merging CityPersons and DhakaAI pseudo-labels; affects label quality and class balance.
  • Augmentation parameters = crop 30-70%, shear ±10°, brightness/exposure ±20%, cutout 15x3%, blur 1px, grayscale 30%, mosaic
    Hand-set augmentation schedule; affects generalization and reported mAP.
  • Camera calibration parameters pi = pi1: 53.9782, 55.5444, 233865.97, 2630553.14; pi2: 60.0909, 46.0172, -78.5563, -76.1593, 234493.41, 2627654.18
    Measured for each site; feed the skew-correction equations for speed and flow. The pi1 list appears to contain only four values for a five-parameter set, which is inconsistent.
  • Line of Interest and speed strip length = LoI user-defined; speed strip 88 ft
    User-selected measurement geometry; flow counts depend on LoI placement and speed estimates depend on strip length.
assumptions (5)
  • domain assumption YOLOv8 pretrained on MS COCO provides transferable features for Bangladeshi traffic objects
    The paper initializes from yolov8s.pt and relies on transfer learning, as described in Model Training.
  • domain assumption Manual post-processing of the field videos yields unbiased ground-truth flow and speed
    Ground truth is collected from the video through manual post-processing in Field Validation; if this reference is biased, the t-test and correlation results inherit the bias.
  • domain assumption Pseudo-labels from separate YOLOv8 models, plus manual correction, produce accurate training labels for the fused dataset
    DhakaPersons labels are generated by applying DhakaAI-trained and CityPersons-trained models to each other's images; the paper states mislabeled annotations are manually corrected but gives no inter-annotator or audit statistics.
  • domain assumption The skew-correction equations (6)-(9) correctly map image coordinates to the ground plane
    Trajectory coordinates are corrected using phi, omega, delta, X0, Y0; the accuracy of speed estimates depends on this projection model.
  • domain assumption DeepSORT's gating and Hungarian association maintain consistent IDs through occlusion
    The paper adopts DeepSORT and acknowledges occlusion causes lost track points; tracking consistency is load-bearing for speed measurement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DEEGITS: Deep Learning based Framework for Measuring Heterogenous Traffic State in Challenging Traffic Scenarios." pith.science (2026). https://pith.science/paper/M5546AFL

@misc{pith2026241108335,
  author       = {Pith},
  title        = {Pith review of: DEEGITS: Deep Learning based Framework for Measuring Heterogenous Traffic State in Challenging Traffic Scenarios},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M5546AFL}},
  note         = {Machine review of arXiv:2411.08335}
}
read the original abstract

This paper presents DEEGITS (Deep Learning Based Heterogeneous Traffic State Measurement), a comprehensive framework that leverages state-of-the-art convolutional neural network (CNN) techniques to accurately and rapidly detect vehicles and pedestrians, as well as to measure traffic states in challenging scenarios (i.e., congestion, occlusion). In this study, we enhance the training dataset through data fusion, enabling simultaneous detection of vehicles and pedestrians. Image preprocessing and augmentation are subsequently performed to improve the quality and quantity of the dataset. Transfer learning is applied on the YOLOv8 pretrained model to increase the model's capability to identify a diverse array of vehicles. Optimal hyperparameters are obtained using the Grid Search algorithm, with the Stochastic Gradient Descent (SGD) optimizer outperforming other optimizers under these settings. Extensive experimentation and evaluation demonstrate substantial accuracy within the detection framework, with the model achieving 0.794 mAP@0.5 on the validation set and 0.786 mAP@0.5 on the test set, surpassing previous benchmarks on similar datasets. The DeepSORT multi-object tracking algorithm is incorporated to track detected vehicles and pedestrians in this study. Finally, the framework is tested to measure heterogeneous traffic states in mixed traffic conditions. Two locations with differing traffic compositions and congestion levels are selected: one motorized-dominant location with moderate density and one non-motorized-dominant location with higher density. Errors are statistically insignificant for both cases, showing correlations from 0.99 to 0.88 and 0.91 to 0.97 for heterogeneous traffic flow and speed measurements, respectively.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 38 canonical work pages

  1. [1]

    Wang, S., A. U. Z. Patwary, W. Huang, and H. K. Lo. A General Framework for Combining Traffic Flow Models and Bayesian Network for Traffic Parameters Estimation. Transportation Research Part C: Emerging Technologies, Vol. 139, 2022, p. 103664

  2. [2]

    Abdel-Aty, M., Z. Wang, O. Zheng, and A. Abdelraouf. Advances and Applications of Computer Vision Techniques in Vehicle Trajectory Generation and Surrogate Traffic Safety Indicators. Accident Analysis & Prevention, Vol. 191, 2023, p. 107191

  3. [3]

    Zhao, J., H. Xu, H. Liu, J. Wu, Y. Zheng, and D. Wu. Detection and Tracking of Pedestrians and Vehicles Using Roadside LiDAR Sensors. Transportation Research Part C: Emerging Technologies, Vol. 100, 2019, pp. 68–87

  4. [4]

    Sahin, and M

    Ahmed, U., O. Sahin, and M. Cetin. Minimizing GPS Dependency for a Vehicle’s Trajectory Identification by Using Data from Smartphone Inertial Sensors and Onboard Diagnostics Device. Transportation Research Record: Journal of the Transportation Research Board, Vol. 2644, No. 1, 2017, pp. 55–63

  5. [5]

    W., H.-C

    Ham, S. W., H.-C. Park, E.-J. Kim, S.-Y. Kho, and D.-K. Kim. Investigating the Influential Factors for Practical Application of Multi-Class Vehicle Detection for Images from Unmanned Aerial Vehicle Using Deep Learning Models. Transportation Research Record: Journal of the Transportation Research Board, Vol. 2674, No. 12, 2020, pp. 553–567

  6. [6]

    Campani, and V

    Giachetti, A., M. Campani, and V. Torre. The Use of Optical Flow for Road Navigation. IEEE Transactions on Robotics and Automation, Vol. 14, No. 1, 1998, pp. 34–48

  7. [7]

    Zangenehpour, S., L. F. Miranda-Moreno, and N. Saunier. Automated Classification Based on Video Data at Intersections with Heavy Pedestrian and Bicycle Traffic: Methodology and Application. Transportation Research Part C: Emerging Technologies, Vol. 56, 2015, pp. 161–176

  8. [8]

    Efficient Feature Selection and Classification for Vehicle Detection

    Xuezhi Wen, Ling Shao, Wei Fang, and Yu Xue. Efficient Feature Selection and Classification for Vehicle Detection. IEEE Transactions on Circuits and Systems for Video Technology, Vol. 25, No. 3, 2015, pp. 508–517

Show all 39 references
  1. [9]

    Yang, Z., and L. S. C. Pun-Cheng. Vehicle Detection in Intelligent Transportation Systems and Its Applications under Varying Environments: A Review. Image and Vision Computing, Vol. 69, 2018, pp. 143–154

  2. [10]

    Marković, K

    Dabiri, S., N. Marković, K. Heaslip, and C. K. Reddy. A Deep Convolutional Neural Network Based Approach for Vehicle Classification Using Large-Scale GPS Trajectory Data. Transportation Research Part C: Emerging Technologies, Vol. 116, 2020, p. 102644

  3. [11]

    Wang, H., Yi. Yu, Y. Cai, X. Chen, L. Chen, and Q. Liu. A Comparative Study of State-of-the-Art Deep Learning Algorithms for Vehicle Detection. IEEE Intelligent Transportation Systems Magazine, Vol. 11, No. 2, 2019, pp. 82–95. Islam, Haque, Hadiuzzaman 20

  4. [12]

    Anguelov, D

    Liu, W., 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 (B. Leibe, J. Matas, N. Sebe, and M. Welling, eds.), Springer International Publishing, Cham, pp. 21–37

  5. [13]

    Divvala, R

    Redmon, J., S. Divvala, R. Girshick, and A. Farhadi. You Only Look Once: Unified, Real-Time Object Detection. Presented at the Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016

  6. [14]

    Pang, and Q

    Tan, M., R. Pang, and Q. V. Le. EfficientDet: Scalable and Efficient Object Detection. Presented at the Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020

  7. [15]

    Duan, K., S. Bai, L. Xie, H. Qi, Q. Huang, and Q. Tian. CenterNet: Keypoint Triplets for Object Detection. Presented at the 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, Korea (South), 2019

  8. [16]

    Brown, and J

    Fan, Q., L. Brown, and J. Smith. A Closer Look at Faster R-CNN for Vehicle Detection. Presented at the 2016 IEEE Intelligent Vehicles Symposium (IV), Gotenburg, Sweden, 2016

  9. [17]

    Spanhel, and A

    Sochor, J., J. Spanhel, and A. Herout. BoxCars: Improving Fine-Grained Recognition of Vehicles Using 3-D Bounding Boxes in Traffic Surveillance. IEEE Transactions on Intelligent Transportation Systems, Vol. 20, No. 1, 2019, pp. 97–108

  10. [18]

    Liang, J., X. Chen, M. He, L. Chen, T. Cai, and N. Zhu. Car Detection and Classification Using Cascade Model. IET Intelligent Transport Systems, Vol. 12, No. 10, 2018, pp. 1201–1209

  11. [19]

    Zarzour, A

    Bouguettaya, A., H. Zarzour, A. Kechida, and A. M. Taberkit. Deep Learning Techniques to Classify Agricultural Crops through UAV Imagery: A Review. Neural Computing and Applications, Vol. 34, No. 12, 2022, pp. 9511–9536

  12. [20]

    Huangfu, L

    Tan, L., T. Huangfu, L. Wu, and W. Chen. Comparison of RetinaNet, SSD, and YOLO v3 for Real- Time Pill Identification. BMC Medical Informatics and Decision Making, Vol. 21, No. 1, 2021, p. 324

  13. [21]

    Zhang, D

    Li, X., Y. Zhang, D. He, X. Teng, B. Liu, and Y. Chen. Passenger Flow Detection in Subway Stations Based on Improved You Only Look Once Algorithm. Transportation Research Record: Journal of the Transportation Research Board, 2023, p. 036119812311591

  14. [22]

    A., and F

    Bin Zuraimi, M. A., and F. H. Kamaru Zaman. Vehicle Detection and Tracking Using YOLO and DeepSORT. Presented at the 2021 IEEE 11th IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE), Penang, Malaysia, 2021

  15. [23]

    Bewley, A., Z. Ge, L. Ott, F. Ramos, and B. Upcroft. Simple Online and Realtime Tracking. Presented at the 2016 IEEE International Conference on Image Processing (ICIP), Phoenix, AZ, USA, 2016

  16. [24]

    Bewley, and D

    Wojke, N., A. Bewley, and D. Paulus. Simple Online and Realtime Tracking with a Deep Association Metric. Presented at the 2017 IEEE International Conference on Image Processing (ICIP), Beijing, 2017

  17. [25]

    Zhang, Y., P. Sun, Y. Jiang, D. Yu, F. Weng, Z. Yuan, P. Luo, W. Liu, and X. Wang. ByteTrack: Multi-Object Tracking by Associating Every Detection Box. In Computer Vision – ECCV 2022, No. 13682, S. Avidan, G. Brostow, M. Cissé, G. M. Farinella, and T. Hassner, eds., Cham, 2022...

  18. [26]

    Lin, T., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. Microsoft COCO: Common Objects in Context. In European Conference on Computer Vision (pp. 740-755). Springer International Publishing, 2014

  19. [27]

    CityPersons: A Diverse Dataset for Pedestrian Detection

    Zhang, S., Benenson, R., & Schiele, B. CityPersons: A Diverse Dataset for Pedestrian Detection. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 3213-3221). IEEE, 2017

  20. [28]

    Stoken, J

    Jocher, G., Ayush Chaurasia, A. Stoken, J. Borovec, NanoCode012, Yonghye Kwon, TaoXie, Kalen Michael, Jiacong Fang, Imyhxy, Lorna, C. Wong, Y. Zeng, Abhiram V, D. Montes, Zhiqiang Wang, C. Fati, Jebastin Nadar, Laughing, UnglvKitDe, Tkianai, YxNONG, P. Skalski, A. Hogan, M. St...

  21. [29]

    Yan, and C

    Dong, X., S. Yan, and C. Duan. A Lightweight Vehicles Detection Network Model Based on YOLOv5. Engineering Applications of Artificial Intelligence, Vol. 113, 2022, p. 104914

  22. [30]

    Kumar, and J

    Sukkar, M., D. Kumar, and J. Sindha. Real-Time Pedestrians Detection by YOLOv5. Presented at the 2021 12th International Conference on Computing Communication and Networking Technologies (ICCCNT), Kharagpur, India, 2021

  23. [31]

    https://github.com/ultralytics/ultralytics

    Ultralytics:YOLOv8. https://github.com/ultralytics/ultralytics. Accessed Jul. 20, 2023

  24. [32]

    Wang, C., Bochkovskiy, A., & Liao, H. Y. M. YOLOv7: Trainable Bag-of-Freebies Sets New State- of-the-Art for Real-Time Object Detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 7464-7475), 2023

  25. [33]

    Shihavuddin, A., & Rashid, M. R. A. (2020). DhakaAI (Version 1.0), Data set. Harvard Dataverse

  26. [34]

    http://www.brta.gov.bd/

    Bangladesh Road Transport Authority (BRTA). http://www.brta.gov.bd/. Accessed Jul. 12, 2023

  27. [35]

    P., & Ba, J

    Kingma, D. P., & Ba, J. Adam: A Method for Stochastic Optimization. In 3rd International Conference for Learning Representations, San Diego, 2015

  28. [36]

    An overview of gradient descent optimization algorithms

    Ruder, S. An overview of gradient descent optimization algorithms. arXiv preprint arXiv:1609.04747, 2016

  29. [37]

    Bin Azad, and Md

    Rahman, R., Z. Bin Azad, and Md. Bakhtiar Hasan. Densely-Populated Traffic Detection Using YOLOv5 and Non-Maximum Suppression Ensembling. In Proceedings of the International Conference on Big Data, IoT, and Machine Learning (M. S. Arefin, M. S. Kaiser, A. Bandyopadhyay, Md. A....

  30. [38]

    Performance Measurement System (PeMS)

    Caltrans. Performance Measurement System (PeMS). https://pems.dot.ca.gov/. Accessed Jul. 12, 2023

  31. [39]

    Rahman, M

    Haque, N., F. Rahman, M. Hadiuzzaman, S. Hossain, M. R. K. Siam, and T. Z. Qiu. PARTS-Based Real-Time Vehicle Detection for Flow Measurement Considering Shadow and Illumination Variation. Presented at the Transportation Research Board 96th Annual Meeting Transportation Researc...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.