An efficient real-time target tracking algorithm using adaptive feature fusion

Liu, Yanyan; Pan, Changcheng; Bie, Minglin; Li, Jin

doi:10.1016/j.jvcir.2022.103505

Cited by 7 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Precision calculations: Precision was a measure of precision that represented the proportion of the example that was actually positive in the example that was divided into positive cases, which was calculated by [2]:…”

Section: Backbonementioning

confidence: 99%

See 1 more Smart Citation

DBF‐YOLO:UAVSmall Targets Detection Based on Shallow Feature Fusion

Liu

Duan

Chen

et al. 2023

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

Driven by deep learning, great breakthroughs had been made in the field of target detection. Small target detection algorithms were widely used in industry, agriculture and other fields. But the small target had few available features and the loss of small target detail information in feature extraction. So it led to the low accuracy of the small target detection algorithms. In this paper, we proposed DBF‐YOLO algorithm based on the classical YOLOV5. The classical YOLOV5 algorithm with high speed. The detection speed of the minimum model could reach 24 ms. However, the deep network structure led to the low detection accuracy of small targets. Our proposed DBF‐YOLO algorithm was an improvement on the problem of small target information being lost. The main contributions of this article were mainly: First, a shallow feature extraction network was introduced in P1 layer, more details of small targets could be well retained. Second, by adding the feature fusion network of shallow feature map and the detection output part in the FPN + PAN layers, the algorithm's accuracy and generalization ability were significantly enhanced. Compared to YOLOV5, the performance of the DBF‐YOLO algorithm was significantly improved. On the validation set, mAP@0.5 and mAP@0.5:0.95 were increased by 8.80 and 5.90%, respectively. Recall was increased from the initial 34.50–41.80%. Precision was increased from initial 44.20 to 50.70%. On the test set, mAP@0.5 and mAP@0.5:0.95 were increased by 6.40 and 3.90%, respectively. Recall was increased 5.10%. Precision was increased 6.60%. Experiments had shown that the improved algorithm achieved good results in accuracy. © 2023 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.

show abstract

Section: Backbonementioning

confidence: 99%

“…As one of the key research areas in artificial intelligence and computer vision, target detection [1] was the cornerstone for solving higher‐level tasks such as image segmentation [1], target tracking [2] and event detection [3]. Nowadays, small target detection [4] still was one of the difficult problems in target detection.…”

Section: Introductionmentioning

confidence: 99%

DBF‐YOLO:UAVSmall Targets Detection Based on Shallow Feature Fusion

Liu

Duan

Chen

et al. 2023

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

show abstract

“…Vision Image BEiT v1 [16], v2 [27], MAE [19], SimMIM [28], ADIOS [29], AMT [30], AttMask [31], Beyond-Masking [32], BootMAE [33], CAE [20], CAN [34], ConvMAE [21], Contrastive MAE [22], ContrastMask [35], dBOT [36], DMAE [37], Denoising MAE [38], GreenMAE [23], iBOT [39], LoMaR [40], LS-MAE [41], MaskAlign [42], MaskDistill [18], MaskFeat [43], MaskTune [44], MetaMask [45], MFM [46], MILAN [47], MixMask [48], MixMIM [24], MRA [49], MSN [50], MST [51], MultiMAE [52], MVP [53], RC-MAE [54], SDMAE [55], SemMAE [56], SdAE [57], SupMAE [58], U-MAE [59], UM-MAE [60] Video AdaMAE [61], Bevt [62], MAM2 [63], MAR [64], MaskViT [65], M3Video [66], MCVD…”

Section: Domain Sub-domain Research Papersmentioning

confidence: 99%

“…The most famous one is Masked Autoencoder (MAE) [19], which owns a very simple learning architecture but has been proven to be a strong and scalable pre-training framework for visual representation learning. MAE has attracted unprecedented attention and got various derivatives (e.g., CAE [20], ConvMAE [21], CMAE [22], GreenMAE [23], MixMIM [24]). To authors' knowledge, no research has ever explored MAE pretrained backbones for scene classification.…”

Section: Introductionmentioning

confidence: 99%

Remote sensing scene classification with masked image modeling

Wang,

Tien

2023

Microwave Remote Sensing: Data Processing and Applications II

View full text Add to dashboard Cite

Remote sensing scene classification has been extensively studied for its critical roles in geological survey, oil exploration, traffic management, earthquake prediction, wildfire monitoring, and intelligence monitoring. In the past, the Machine Learning (ML) methods for performing the task mainly used the backbones pretrained in the manner of supervised learning (SL). As Masked Image Modeling (MIM), a self-supervised learning (SSL) technique, has been shown as a better way for learning visual feature representation, it presents a new opportunity for improving ML performance on the scene classification task. This research aims to explore the potential of MIM pretrained backbones on four well-known classification datasets: Merced, AID, NWPU-RESISC45, and Optimal-31. Compared to the published benchmarks, we show that the MIM pretrained Vision Transformer (ViTs) backbones outperform other alternatives (up to 18% on top 1 accuracy) and that the MIM technique can learn better feature representation than the supervised learning counterparts (up to 5% on top 1 accuracy). Moreover, we show that the general-purpose MIM-pretrained ViTs can achieve competitive performance as the specially designed yet complicated Transformer for Remote Sensing (TRS) framework. Our experiment results also provide a performance baseline for future studies.

show abstract

Multi-template global re-detection based on Gumbel-Softmax in long-term visual tracking

Hou

et al. 2023

Appl Intell

View full text Add to dashboard Cite

An efficient real-time target tracking algorithm using adaptive feature fusion

Cited by 7 publications

References 37 publications

DBF‐YOLO:UAVSmall Targets Detection Based on Shallow Feature Fusion

DBF‐YOLO:UAVSmall Targets Detection Based on Shallow Feature Fusion

Remote sensing scene classification with masked image modeling

Multi-template global re-detection based on Gumbel-Softmax in long-term visual tracking

Contact Info

Product

Resources

About