Active detection for fish species recognition in underwater environments DOI
Chiranjibi Shah, M M Nabi, Simegnew Yihunie Alaba

et al.

Published: June 6, 2024

Fish species must be identified for stock assessments, ecosystem monitoring, production management, and the conservation of endangered species. Implementing algorithms fish detection in underwater settings like Gulf Mexico poses a formidable challenge. Active learning, method that efficiently identifies informative samples annotation while staying within budget, has demonstrated its effectiveness context object recent times. In this study, we present an active model designed recognition environments. This can employed as system to effectively lower expense associated with manual annotation. It uses epistemic uncertainty Evidential Deep Learning (EDL) proposes novel module denoted Model Evidence Head (MEH) employs Hierarchical Uncertainty Aggregation (HUA) obtain informativeness image. We conducted experiments using fine-grained extensive dataset reef collected from Mexico, specifically Southeast Area Monitoring Assessment Program Dataset 2021 (SEAMAPD21). The experimental results demonstrate framework achieves better performance on SEAMAPD21 demonstrating favorable balance between data efficiency recognition.

Language: Английский

A Survey on Deep-Learning-Based LiDAR 3D Object Detection for Autonomous Driving DOI Creative Commons
Simegnew Yihunie Alaba, John Ball

Sensors, Journal Year: 2022, Volume and Issue: 22(24), P. 9577 - 9577

Published: Dec. 7, 2022

LiDAR is a commonly used sensor for autonomous driving to make accurate, robust, and fast decision-making when driving. The in the perception system, especially object detection, understand environment. Although 2D detection has succeeded during deep-learning era, lack of depth information limits understanding environment location. Three-dimensional sensors, such as LiDAR, give 3D about surrounding environment, which essential system. Despite attention computer vision community due multiple applications robotics driving, there are challenges, scale change, sparsity, uneven distribution data, occlusions. Different representations data methods minimize effect sparsity have been proposed. This survey presents LiDAR-based feature-extraction techniques data. coordinate systems differ camera datasets methods. Therefore, summarized. Then, state-of-the-art object-detection reviewed with selected comparison among

Language: Английский

Citations

57
Deep Learning-Based Image 3-D Object Detection for Autonomous Driving: Review DOI
Simegnew Yihunie Alaba, John Ball

IEEE Sensors Journal, Journal Year: 2023, Volume and Issue: 23(4), P. 3378 - 3394

Published: Jan. 13, 2023

An accurate and robust perception system is key to understanding the driving environment of autonomous robots. Autonomous needs 3-D information about objects, including object’s location pose, understand clearly. A camera sensor widely used in because its richness color texture, low price. The major problem with lack information, which necessary environment. In addition, scale change occlusion make object detection more challenging. Many deep learning-based methods, such as depth estimation, have been developed solve information. This survey presents image bounding box encoding techniques evaluation metrics. image-based methods are categorized based on technique estimate an image’s insights added each method. Then, state-of-the-art (SOTA) monocular stereo camera-based summarized. We also compare performance selected models present challenges future directions detection.

Language: Английский

Citations

42
Modeling the satellite instrument visibility range for detecting underwater targets DOI
Jun Chen,

Wenting Quan,

Xianqiang He

et al.

ISPRS Journal of Photogrammetry and Remote Sensing, Journal Year: 2025, Volume and Issue: 222, P. 64 - 78

Published: Feb. 26, 2025

Language: Английский

Citations

1
Off-Road Detection Analysis for Autonomous Ground Vehicles: A Review DOI Creative Commons
Fahmida Islam, M M Nabi, John Ball

et al.

Sensors, Journal Year: 2022, Volume and Issue: 22(21), P. 8463 - 8463

Published: Nov. 3, 2022

When it comes to some essential abilities of autonomous ground vehicles (AGV), detection is one them. In order safely navigate through any known or unknown environment, AGV must be able detect important elements on the path. Detection applicable both on-road and off-road, but they are much different in each environment. The key environment that identify drivable pathway whether there obstacles around it. Many works have been published focusing components various ways. this paper, a survey most recent advancements methods intended specifically for off-road has presented. For this, we divided literature into three major groups: positive negative obstacles. Each portion further multiple categories based technology used, example, single sensor-based, how data analyzed. Furthermore, added critical findings technology, challenges associated with possible future directions. Authors believe work will help reader finding who doing similar works.

Language: Английский

Citations

29
Tackling class imbalance in computer vision: a contemporary review DOI
Manisha Saini, Seba Susan

Artificial Intelligence Review, Journal Year: 2023, Volume and Issue: 56(S1), P. 1279 - 1335

Published: July 20, 2023

Language: Английский

Citations

21
Transferable Deep Learning Model for the Identification of Fish Species for Various Fishing Grounds DOI Creative Commons
Tatsuhito Hasegawa, Kei Kondo,

Hiroshi Senou

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(3), P. 415 - 415

Published: Feb. 26, 2024

The digitization of catch information for the promotion sustainable fisheries is gaining momentum globally. However, manual measurement fundamental information, such as species identification, length measurement, and fish count, highly inconvenient, thus intensifying call its automation. Recently, image recognition systems based on convolutional neural networks (CNNs) have been extensively studied across diverse fields. Nevertheless, deployment CNNs identifying difficult owing to intricate nature managing a plethora species, which fluctuate season locale, in addition scarcity public datasets encompassing large catches. To overcome this issue, we designed transferable pre-trained CNN model specifically can be easily reused various fishing grounds. Utilizing an extensive photographic database from Japanese museum, developed identification (TFI) employing strategies multiple pre-training, learning rate scheduling, multi-task learning, metric learning. We further introduced two application methods, namely transfer output layer masking, TFI model, validating efficacy through rigorous experiments.

Language: Английский

Citations

5
DyFish-DETR: Underwater Fish Image Recognition Based on Detection Transformer DOI Creative Commons
Zhuowei Wang,

Zhukang Ruan,

Chong Chen

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(6), P. 864 - 864

Published: May 22, 2024

Due to the complexity of underwater environments and lack training samples, application target detection algorithms environment has yet provide satisfactory results. It is crucial design specialized recognition for different tasks. In order achieve this goal, we created a dataset freshwater fish captured from multiple angles lighting conditions, aiming improve in natural environments. We propose method suitable detection, called DyFish-DETR (Dynamic Fish Detection with Transformers). DyFish-DETR, DyFishNet Net) better extract body texture features. A Slim Hybrid Encoder designed fuse feature information. The results ablation experiments show that can effectively mean Average Precision (mAP) model detection. Frame Per Second (FPS). Both reduce parameters Floating Point Operations (FLOPs). our proposed dataset, achieved mAP 96.6%. benchmarking experimental (AP) Recall (AR) are higher than several state-of-the-art methods. Additionally, respectively, 99%, 98.8%, 83.2% other datasets.

Language: Английский

Citations

5
Object detection in power line infrastructure: A review of the challenges and solutions DOI
Pratibha Sharma, Sumeet Saurav, Sanjay Kumar Singh

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2023, Volume and Issue: 130, P. 107781 - 107781

Published: Dec. 27, 2023

Language: Английский

Citations

13
DiffusionFR: Species Recognition of Fish in Blurry Scenarios via Diffusion and Attention DOI Creative Commons
Guoying Wang, Bing Shi, Xiaomei Yi

et al.

Animals, Journal Year: 2024, Volume and Issue: 14(3), P. 499 - 499

Published: Feb. 2, 2024

Blurry scenarios, such as light reflections and water ripples, often affect the clarity signal-to-noise ratio of fish images, posing significant challenges for traditional deep learning models in accurately recognizing species. Firstly, rely on a large amount labeled data. However, it is difficult to label data blurry scenarios. Secondly, existing need be more effective processing bad, blurry, otherwise inadequate which an essential reason their low recognition rate. A method based diffusion model attention mechanism image DiffusionFR, proposed solve these problems improve performance species images This paper presents selection application this correcting technique. In method, two-stage network model, TSD, designed deblur scene pictures restore clarity, learnable module, LAM, intended accuracy recognition. addition, new dataset BlurryFish, was constructed used validate effectiveness combining from publicly available Fish4Knowledge. The experimental results demonstrate that DiffusionFR achieves outstanding various datasets. On original dataset, achieved highest training 97.55%, well Top-1 test score 92.02% Top-5 95.17%. Furthermore, nine datasets with reflection noise, mean values reached peak at 96.50%, while were 90.96% 94.12%, respectively. Similarly, three ripple 95.00%, 89.54% 92.73%, These showcases superior enhanced robustness handling noise.

Language: Английский

Citations

4
Triple Attention Mechanism with YOLOv5s for Fish Detection DOI Creative Commons
Wei Long, Yawen Wang, Lingxi Hu

et al.

Fishes, Journal Year: 2024, Volume and Issue: 9(5), P. 151 - 151

Published: April 23, 2024

Traditional fish farming methods suffer from backward production, low efficiency, yield, and environmental pollution. As a result of thorough research using deep learning technology, the industrial aquaculture model has experienced gradual maturation. A variety complex factors makes it difficult to extract effective features, which results in less-than-good performance. This paper proposes detection method that combines triple attention mechanism with You Only Look Once (TAM-YOLO)model. In order enhance speed training, process data encapsulation incorporates positive sample matching. An exponential moving average (EMA) is incorporated into training make more robust, coordinate (CA) convolutional block module are integrated YOLOv5s backbone feature extraction channels spatial locations. The extracted maps input PANet path aggregation network, underlying information stacked maps. improves accuracy underwater blurred distorted images. Experimental show proposed TAM-YOLO outperforms YOLOv3, YOLOv4, YOLOv5s, YOLOv5m, SSD, mAP value 95.88%, thus providing new strategy for detection.

Language: Английский

Citations

4