Dataset: Traffic Images Captured from UAVs for Use in Training Machine Vision Algorithms for Traffic Management

Rosende, Sergio Bemposta; Ghisler, Sergio; Andrés, Javier Fernández; Sánchez-Soriano, Javier

doi:10.3390/data7050053

Cited by 5 publications

(6 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regarding data, two datasets of aerial images were used, both developed by the Intelligent Control Systems (SIC) group at UEM: (1) "Traffic Images Captured from UAVs for Use in Training Machine Vision Algorithms for Traffic Management" [23] and (2) "Roundabout Aerial Images for Vehicle Detection" [24].…”

Section: Datasetsmentioning

confidence: 99%

“…Two datasets created by the Intelligent Control Systems (SIC) Research Group at the European University of Madrid (UEM) were used and combined to create a new one. The first dataset, generated using the CVAT annotation tool [38], was documented in the article titled "Traffic Images Captured from UAVs for Use in Training Machine Vision Algorithms for Traffic Management" [23]. The second dataset, "Roundabout Aerial Images for Vehicle Detection" [24], was annotated using the PASCAL VOC XML technique, unlike the first dataset, which used YOLO annotation.…”

Section: Dataset Generationmentioning

confidence: 99%

“…The data presented in this study are openly available in https:// zenodo.org/record/5776219 (accessed on 1 November 2023) with doi: https://doi.org/10.3390/ data7050053 [23] and https://zenodo.org/records/6407460 (accessed on 1 November 2023) with doi: https://doi.org/10.3390/data7040047 [24].…”

Section: Data Availability Statementmentioning

confidence: 99%

“…To meet these requirements, this work utilizes two different datasets that provide a substantial amount of data, a diverse range of scenarios, and recognition of numerous objects within each image [23,24].…”

mentioning

confidence: 99%

See 3 more Smart Citations

Implementation of an Edge-Computing Vision System on Reduced-Board Computers Embedded in UAVs for Intelligent Traffic Management

Bemposta Rosende,

Ghisler,

Fernández-Andrés

et al. 2023

Drones

Self Cite

View full text Add to dashboard Cite

Advancements in autonomous driving have seen unprecedented improvement in recent years. This work addresses the challenge of enhancing the navigation of autonomous vehicles in complex urban environments such as intersections and roundabouts through the integration of computer vision and unmanned aerial vehicles (UAVs). UAVs, owing to their aerial perspective, offer a more effective means of detecting vehicles involved in these maneuvers. The primary objective is to develop, evaluate, and compare different computer vision models and reduced-board (and small-power) hardware for optimizing traffic management in these scenarios. A dataset was constructed using two sources, several models (YOLO 5 and 8, DETR, and EfficientDetLite) were selected and trained, four reduced-board computers were chosen (Raspberry Pi 3B+ and 4, Jetson Nano, and Google Coral), and the models were tested on these boards for edge computing in UAVs. The experiments considered training times (with the dataset and its optimized version), model metrics were obtained, inference frames per second (FPS) were measured, and energy consumption was quantified. After the experiments, it was observed that the combination that best suits our use case is the YoloV8 model with the Jetson Nano. On the other hand, a combination with much higher inference speed but lower accuracy involves the EfficientDetLite models with the Google Coral board.

show abstract

Section: Datasetsmentioning

confidence: 99%

Section: Dataset Generationmentioning

confidence: 99%

Section: Data Availability Statementmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Implementation of an Edge-Computing Vision System on Reduced-Board Computers Embedded in UAVs for Intelligent Traffic Management

Bemposta Rosende,

Ghisler,

Fernández-Andrés

et al. 2023

Drones

Self Cite

View full text Add to dashboard Cite

show abstract

“…This may cause the network to learn to always expect to detect objects in the center of the region or image. This will not be a problem for this project, given that we have used YOLO as a CNN, which initially divides the image into sections (by default, into 7x7 sections) [19] and that each section has the same probability of containing an object regardless of its position [20,21].…”

Section: Figure 3 Application With Blurred Targets (White Box)mentioning

confidence: 99%

Optimization Algorithm to Reduce Training Time for Deep Learning Computer Vision Algorithms Using Large Image Datasets With Tiny Objects

Rosende,

Fernández-Andrés,

Sánchez-Soriano

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

The optimization of convolutional neural networks (CNN) generally refers to the improvement of the inference process, making this as fast and precise as possible. While inference time is an essential factor in using these networks in real time, the training of CNNs using very large datasets can be very costly in terms of time and computing power. This paper proposes a technique to reduce training time by an average of 75% without altering the results of CNN training with an algorithm which partitions the dataset and discards superfluous objects (targets). This algorithm is a tool that pre-processes the original dataset, generating a smaller and more condensed dataset to be used for network training. The effectiveness of this tool depends on the type of dataset used for training the CNN and is particularly effective with sequential images (video), large images and images with tiny targets generally from drones or traffic surveillance cameras (but applicable to any other type of image which meets the requirements). The tool can be parameterized to meet the characteristics of the initial dataset.

show abstract

Detection and identification drones using long short-term memory and Bayesian optimization

El-Latif

2024

Multimed Tools Appl

View full text Add to dashboard Cite

This paper proposed a model based on bidirectional Long Short-Term Memory (Bi-LSTM) and Bayesian optimization to detect different drones in different Scenarios. Six different drones in three distinct scenarios—cloudy, sunny, and evening—are included in the dataset. The three primary stages of the suggested model are pre-processing, hyper-parameter optimization, and classification phase. Images are resized, noise is reduced, and contrast is enhanced in the first step. The hyperparameter values are then chosen in the second step using Bayesian optimization. In the end, the proposed model is constructed with nine layers based on the Gated Recurrent Unit (GRU) and Bi-LSTM for classification. For the cloudy scenario, the model achieves 97.43% accuracy, 99.52% sensitivity, 92.79% specificity, 96.64% precision, and 98.06 F1-score. In addition, the sunny scenario achieves 93.87%in accuracy, 97.14% in sensitivity, 87.06% in specificity, 94% in precision, and 90.25 in F1_score. The final experiment in the evening scenario is 97.50% accuracy, 99.39% sensitivity, 93.90% specificity, and 96.89% precision. Comparative results are presented at the end of the paper, and it shows that the proposed model overcomes previous works that used the same dataset.

show abstract

Dataset: Traffic Images Captured from UAVs for Use in Training Machine Vision Algorithms for Traffic Management

Cited by 5 publications

References 25 publications

Implementation of an Edge-Computing Vision System on Reduced-Board Computers Embedded in UAVs for Intelligent Traffic Management

Implementation of an Edge-Computing Vision System on Reduced-Board Computers Embedded in UAVs for Intelligent Traffic Management

Optimization Algorithm to Reduce Training Time for Deep Learning Computer Vision Algorithms Using Large Image Datasets With Tiny Objects

Detection and identification drones using long short-term memory and Bayesian optimization

Contact Info

Product

Resources

About