LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost
Efficient deep learning models are crucial in resource-constrained environments, especially for marine image classification in underwater monitoring and biodiversity assessment. This paper presents LatentResNet, a computationally lightweight deep learning model involving two key innovations: (i) usi...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/13/6/1019 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839653635889299456 |
|---|---|
| author | Muhab Hariri Ercan Avsar Ahmet Aydın |
| author_facet | Muhab Hariri Ercan Avsar Ahmet Aydın |
| author_sort | Muhab Hariri |
| collection | DOAJ |
| description | Efficient deep learning models are crucial in resource-constrained environments, especially for marine image classification in underwater monitoring and biodiversity assessment. This paper presents LatentResNet, a computationally lightweight deep learning model involving two key innovations: (i) using the encoder from the proposed LiteAE, a lightweight autoencoder for image reconstruction, as input to the model to reduce the spatial dimension of the data and (ii) integrating a DeepResNet architecture with lightweight feature extraction components to refine encoder-extracted features. LiteAE demonstrated high-quality image reconstruction within a single training epoch. LatentResNet variants (large, medium, and small) are evaluated on ImageNet-1K to assess their efficiency against state-of-the-art models and on Fish4Knowledge for domain-specific performance. On ImageNet-1K, the large variant achieves 66.3% top-1 accuracy (1.7M parameters, 0.2 GFLOPs). The medium and small variants reach 60.8% (1M, 0.1 GFLOPs) and 54.8% (0.7M, 0.06 GFLOPs), respectively. After fine-tuning on Fish4Knowledge, the large, medium, and small variants achieve 99.7%, 99.8%, and 99.7%, respectively, outperforming the classification metrics of benchmark models trained on the same dataset, with up to 97.4% and 92.8% reductions in parameters and FLOPs, respectively. The results demonstrate LatentResNet’s effectiveness as a lightweight solution for real-world marine applications, offering accurate and lightweight underwater vision. |
| format | Article |
| id | doaj-art-a55a74ba4c304b45aea95a1a1e621a5f |
| institution | Matheson Library |
| issn | 2077-1312 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj-art-a55a74ba4c304b45aea95a1a1e621a5f2025-06-25T14:01:04ZengMDPI AGJournal of Marine Science and Engineering2077-13122025-05-01136101910.3390/jmse13061019LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational CostMuhab Hariri0Ercan Avsar1Ahmet Aydın2Electrical and Electronics Engineering Department, Çukurova University, 01330 Adana, TurkeySection for Fisheries Technology, National Institute of Aquatic Resources (DTU Aqua), Technical University of Denmark, 9850 Hirtshals, DenmarkBiomedical Engineering Department, Çukurova University, 01330 Adana, TurkeyEfficient deep learning models are crucial in resource-constrained environments, especially for marine image classification in underwater monitoring and biodiversity assessment. This paper presents LatentResNet, a computationally lightweight deep learning model involving two key innovations: (i) using the encoder from the proposed LiteAE, a lightweight autoencoder for image reconstruction, as input to the model to reduce the spatial dimension of the data and (ii) integrating a DeepResNet architecture with lightweight feature extraction components to refine encoder-extracted features. LiteAE demonstrated high-quality image reconstruction within a single training epoch. LatentResNet variants (large, medium, and small) are evaluated on ImageNet-1K to assess their efficiency against state-of-the-art models and on Fish4Knowledge for domain-specific performance. On ImageNet-1K, the large variant achieves 66.3% top-1 accuracy (1.7M parameters, 0.2 GFLOPs). The medium and small variants reach 60.8% (1M, 0.1 GFLOPs) and 54.8% (0.7M, 0.06 GFLOPs), respectively. After fine-tuning on Fish4Knowledge, the large, medium, and small variants achieve 99.7%, 99.8%, and 99.7%, respectively, outperforming the classification metrics of benchmark models trained on the same dataset, with up to 97.4% and 92.8% reductions in parameters and FLOPs, respectively. The results demonstrate LatentResNet’s effectiveness as a lightweight solution for real-world marine applications, offering accurate and lightweight underwater vision.https://www.mdpi.com/2077-1312/13/6/1019lightweight neural networksdeep learningunderwater image classificationcomputational efficiencyautoencoderreal-time processing |
| spellingShingle | Muhab Hariri Ercan Avsar Ahmet Aydın LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost Journal of Marine Science and Engineering lightweight neural networks deep learning underwater image classification computational efficiency autoencoder real-time processing |
| title | LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost |
| title_full | LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost |
| title_fullStr | LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost |
| title_full_unstemmed | LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost |
| title_short | LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost |
| title_sort | latentresnet an optimized underwater fish classification model with a low computational cost |
| topic | lightweight neural networks deep learning underwater image classification computational efficiency autoencoder real-time processing |
| url | https://www.mdpi.com/2077-1312/13/6/1019 |
| work_keys_str_mv | AT muhabhariri latentresnetanoptimizedunderwaterfishclassificationmodelwithalowcomputationalcost AT ercanavsar latentresnetanoptimizedunderwaterfishclassificationmodelwithalowcomputationalcost AT ahmetaydın latentresnetanoptimizedunderwaterfishclassificationmodelwithalowcomputationalcost |