Lightweight and computationally faster Hypermetropic Convolutional Neural Network for small size object detection

Amudhan, A. N.; Sudheer, A. P.

doi:10.1016/j.imavis.2022.104396

Cited by 24 publications

(10 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Liu Y et al [29] introduced the Multi-Branch Parallel Feature Pyramid Network (MPFPN), which adds two additional parallel branches to FPN, therefore enhancing the network's capability to extract feature information for small targets. Amudhan et al [30] considered the supportive role of contextual information for small objects, and built skip connections between shallow and deep features, effectively enhancing the performance of small object detection in aerial images.…”

Section: Multiscale Feature Fusionmentioning

confidence: 99%

GA-Net: Accurate and Efficient Object Detection on UAV Images Based on Grid Activations

Zhang,

Luo,

et al. 2024

Drones

View full text Add to dashboard Cite

Object detection plays a crucial role in unmanned aerial vehicle (UAV) missions, where captured objects are often small and require high-resolution processing. However, this requirement is always in conflict with limited computing resources, vast fields of view, and low latency requirements. To tackle these issues, we propose GA-Net, a novel approach tailored for UAV images. The key innovation includes the Grid Activation Module (GAM), which efficiently calculates grid activations, the probability of foreground presence at grid scale. With grid activations, the GAM helps filter out patches without objects, minimize redundant computations, and improve inference speeds. Additionally, the Grid-based Dynamic Sample Selection (GDSS) focuses the model on discriminating positive samples and hard negatives, addressing background bias during training. Further enhancements involve GhostFPN, which refines Feature Pyramid Network (FPN) using Ghost module and depth-wise separable convolution. This not only expands the receptive field for improved accuracy, but also reduces computational complexity. We conducted comprehensive evaluations on DGTA-Cattle-v2, a synthetic dataset with added background images, and three public datasets (VisDrone, SeaDronesSee, DOTA) from diverse domains. The results prove the effectiveness and practical applicability of GA-Net. Despite the common accuracy and speed trade-off challenge, our GA-Net successfully achieves a mutually beneficial scenario through the strategic use of grid activations.

show abstract

Section: Multiscale Feature Fusionmentioning

confidence: 99%

GA-Net: Accurate and Efficient Object Detection on UAV Images Based on Grid Activations

Zhang,

Luo,

et al. 2024

Drones

View full text Add to dashboard Cite

show abstract

“…Through encoding and decoding operations, this network densely concatenates spatial and semantic information to extract features of small objects. Amudhan et al 13 developed the CNN, which improves the detection efficiency of small objects by extracting more features from shallow layers and transferring the low-level features to deeper layers. Besides, researchers can expand the number of small objects by processing the data, increasing the input resolution of the data, or increasing the samples of small object.…”

Section: Small Object Detectionmentioning

confidence: 99%

“…Through encoding and decoding operations, this network densely concatenates spatial and semantic information to extract features of small objects. Amudhan et al 13 . developed the CNN, which improves the detection efficiency of small objects by extracting more features from shallow layers and transferring the low-level features to deeper layers.…”

Section: Related Workmentioning

confidence: 99%

Small object in the dark light scene using dual-branch channel and spatial

Xia

Shi

2023

J. Electron. Imag.

View full text Add to dashboard Cite

.Object detection is an enabling technology of computer vision for locating instances and categorizing classes of objects in images or videos. It has made enormous strides with significant growing of deep learning research over the last decade. However, there are severe challenges on small object detection in complex scenes as they appear with low resolution and have not enough contrast from the background information. This disturbance may cause missed detection and detection accuracy decline for small objects. We propose a small object detection network in the dark light scene based on improved YOLOv5. The network takes YOLOv5 as the baseline and incorporates a channel and spatial dual-branch backbone module, which enhances the details by fusing the features of the two branches. We also introduce a densely linked feature fusion network before detection layers with receptive field block. The fusion network integrates deep features with shallow ones across different scales. A data augmentation module is used to enhance the brightness and limit the contrast for small object detection in the dark light scene. Experiments are carried out on the Dark Face and darkened Vis Drone dataset. The results show that the evaluation index of the proposed method is better than that of the comparison methods. From the detected images, it is obvious that the undetected frame of the small object in the dark light scene decreases and the detection accuracy improves. All of the results show that our model has better performance than some existing methods for small object detection in the dark light scene.

show abstract

“…In [ 10 ], both Faster-RCNN and YOLO were built and exhibited comparable detection accuracy; however, YOLO requires shorter execution time because to the one-stage architecture’s benefit [ 33 ]. CNN-based deep learning models have been described in [ 1 , 21 , 38 , 54 ], and [ 14 ]. The merging of CNN and LSTM for activity identification was studied, for example, in [ 21 , 54 ].…”

Section: Related Workmentioning

confidence: 99%

“… 83% [ 34 ] 2020 Ren et al YOLO with SqueezeNet PASCAL VOC dataset. 72% [ 30 ] 2022 Qi et al YOLO It was tested using 300 field images 92% [ 38 ] 2022 Sha et al CNN-based approach CityPersons and the ETH dataset – [ 13 ] 2022 Intasuttisak et al Yolo-V5 125 images captured by a drone-based camera 92% [ 49 ] 2022 Wu et al YOLO with Im-Res2Net-101 Public dataset from Kaggle 92% [ 1 ] 2022 Amudhan et al Hypermetropic CNN approach VEDAI and Visdrone dataset 61% [ 14 ] 2021 Junos et al Lightweight CNN model VEDAI dataset 47% [ 44 ] 2022 Triki et al Improved YOLO-V3 Specimen dataset with 4000 images 95% [ 10 ] 2020 Hung et al Faster R-CNN They used UAV123 public benchmark for pedestrian identification and drone-captured photos. 98% [ 4 ] 2023 Bao et al YOLOv5 They employed 350 UAVs, each of which took a JPEG photograph 8000 by 6000 pixels in size.…”

Section: Related Workmentioning

confidence: 99%

Optimized deep learning vision system for human action recognition from drone images

Samma

Sama

2023

Multimed Tools Appl

View full text Add to dashboard Cite

There are several benefits to constructing a lightweight vision system that is implemented directly on limited hardware devices. Most deep learning-based computer vision systems, such as YOLO (You Only Look Once), use computationally expensive backbone feature extractor networks, such as ResNet and Inception network. To address the issue of network complexity, researchers created SqueezeNet, an alternative compressed and diminutive network. However, SqueezeNet was trained to recognize 1000 unique objects as a broad classification system. This work integrates a two-layer particle swarm optimizer (TLPSO) into YOLO to reduce the contribution of SqueezeNet convolutional filters that have contributed less to human action recognition. In short, this work introduces a lightweight vision system with an optimized SqueezeNet backbone feature extraction network. Secondly, it does so without sacrificing accuracy. This is because that the high-dimensional SqueezeNet convolutional filter selection is supported by the efficient TLPSO algorithm. The proposed vision system has been used to the recognition of human behaviors from drone-mounted camera images. This study focused on two separate motions, namely walking and running. As a consequence, a total of 300 pictures were taken at various places, angles, and weather conditions, with 100 shots capturing running and 200 images capturing walking. The TLPSO technique lowered SqueezeNet’s convolutional filters by 52%, resulting in a sevenfold boost in detection speed. With an F1 score of 94.65% and an inference time of 0.061 milliseconds, the suggested system beat earlier vision systems in terms of human recognition from drone-based photographs. In addition, the performance assessment of TLPSO in comparison to other related optimizers found that TLPSO had a better convergence curve and achieved a higher fitness value. In statistical comparisons, TLPSO surpassed PSO and RLMPSO by a wide margin.

show abstract

Lightweight and computationally faster Hypermetropic Convolutional Neural Network for small size object detection

Cited by 24 publications

References 25 publications

GA-Net: Accurate and Efficient Object Detection on UAV Images Based on Grid Activations

GA-Net: Accurate and Efficient Object Detection on UAV Images Based on Grid Activations

Small object in the dark light scene using dual-branch channel and spatial

Optimized deep learning vision system for human action recognition from drone images

Contact Info

Product

Resources

About