Probability-Based Channel Pruning for Depthwise Separable Convolutional Networks

Zhao, Hanli; Shi, Kaijie; Jin, Xiaogang; Xu, Mingliang; Huang, Hui; Lu, Wanglong; Liu, Ying

doi:10.1007/s11390-022-2131-8

Cited by 4 publications

(5 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…13 Deep NNs have demonstrated remarkable performance on a wide range of visual tasks, such as object detection, 14,15 object tracking, 16,17 segmentation, 18,19 and image classification. 13,[20][21][22] Deep convolutional neural networks (CNN) [23][24][25][26] and transformers [27][28][29][30][31] are two good examples. To recognize images accurately, researchers have proposed various architectures and techniques for CNNs, such as using multiple layers, 23 skip connections, 24 dense connections, 25 squeeze and excitation steps, 32 attention mechanisms, 33 and large kernel attention.…”

Section: Image Recognition Techniquesmentioning

confidence: 99%

See 1 more Smart Citation

TEG: image theme recognition using text-embedding-guided few-shot adaptation

Wang,

Lu,

Wang

et al. 2024

J. Electron. Imag.

View full text Add to dashboard Cite

Grouping images into different themes is a challenging task in photo book curation. Unlike image object recognition, image theme recognition focuses on the understanding of the main subject or overall meaning conveyed by an image. However, it is challenging to achieve satisfactory performance using existing general image recognition methods. In this work, we aim to solve the image theme recognition task with few-shot training samples using pre-trained contrastive language-image models. A text-prompt-guided few-shot image adaptation framework is proposed, which incorporates a text-embedding-guided classifier and an auxiliary classification loss to exploit embedded visual and text features, stabilize the network training, and enhance recognition performance. We also present an annotated dataset Theme25 for studying image theme recognition. We conducted experiments on our Theme25 dataset as well as the publicly available CIFAR100 and ImageNet datasets to demonstrate the superiority of our method over the compared stateof-the-art methods.

show abstract

Section: Image Recognition Techniquesmentioning

confidence: 99%

“…Deep NNs have demonstrated remarkable performance on a wide range of visual tasks, such as object detection, 14 , 15 object tracking, 16 , 17 segmentation, 18 , 19 and image classification 13 , 20 – 22 Deep convolutional neural networks (CNN) 23 – 26 and transformers 27 – 31 are two good examples.…”

Section: Related Workmentioning

confidence: 99%

TEG: image theme recognition using text-embedding-guided few-shot adaptation

Wang,

Lu,

Wang

et al. 2024

J. Electron. Imag.

View full text Add to dashboard Cite

show abstract

“…In the original YOLOv3 algorithm, the K-means algorithm was used to cluster the labeled boxes in the VOC and COCO datasets, and nine anchor boxes with sizes of (10, 13), (16,30), (33,23), (30,61), (62, 45), (59, 119), (116, 90), (156,198), and (373, 326), were used. The VOC and COCO datasets included multiple types of targets.…”

Section: K-means Clustering Of Uav Datasetsmentioning

confidence: 99%

“…With K = 9, the clustered results of the K-means algorithm on the UAV dataset labeled boxes are shown in Figure 2, and the sizes of the obtained nine new anchor boxes were (8,18), (10,30), (17,14), (31,32), (16,40), (15,26), (23,47), (31,70), and (102, 102).…”

Section: K-means Clustering Of Uav Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Detection of Drones Using Channel and Layer Pruning, Based on the YOLOv3-SPP3 Deep Learning Algorithm

2022

View full text Add to dashboard Cite

Achieving a real-time and accurate detection of drones in natural environments is essential for the interception of drones intruding into high-security areas. However, a rapid and accurate detection of drones is difficult because of their small size and fast speed. In this paper a drone detection method as proposed by pruning the convolutional channel and residual structures of YOLOv3-SPP3. First, the k-means algorithm was used to cluster label the boxes. Second, the channel and shortcut layer pruning algorithm was used to prune the model. Third, the model was fine tuned to achieve a real-time detection of drones. The experimental results obtained by using the Ubuntu server under the Python 3.6 environment show that the YOLOv3-SPP3 algorithm is better than YOLOV3, Tiny-YOLOv3, CenterNet, SSD300, and faster R-CNN. There is significant compression in the size, the maximum compression factor is 20.1 times, the maximum detection speed is increased by 10.2 times, the maximum map value is increased by 15.2%, and the maximum precision is increased by 16.54%. The proposed algorithm achieves the mAP score of 95.15% and the detection speed of 112 f/s, which can meet the requirements of the real-time detection of UAVs.

show abstract