Deep Learning for Marine Science

scroll to explore
Cover image for research topic "Deep Learning for Marine Science"
ChatGPT summary

Harnessing the power of big data for detection, classification and segmentation of objects in the ocean

Underwater research is a rapidly growing field, with new methods being developed to address a variety of challenges. This research topic covers a range of topics, including:

  • Sound speed profile inversion using task-driven meta-deep-learning (TDML) frameworks
  • Multi-modal data fusion for mesoscale eddy detection
  • Intelligent acoustic tracking models for cetacean conservation
  • Underwater optical communication (UWOC) channel emulation
  • Deep learning models for seagrass detection
  • Object detection for suspended particle abundance
  • Generative adversarial networks (GANs) for underwater monocular SLAM
  • Simultaneous restoration and super-resolution GANs (SRSRGANs) for image quality improvement
  • Deep learning models for Southwestern Atlantic Front (SAF) detection
  • Adaptive sampling for marine plankton using edge servers and data visualization
  • Automatic detection and classification of echo traces of Pacific saury
  • Image-based machine learning methods for data analysis
  • Deep learning algorithms for fish population assessment
  • Underwater image restoration technology
  • Weakly supervised learning for marine life data labeling
  • Multi-scale fusion methods for underwater image contrast improvement
  • Random forest algorithms for pCO2 modeling
  • Convolutional neural networks (CNNs) for ship classification
  • YOLOv5 algorithms for underwater target detection
  • Transformer-based frameworks for marine fish image classification
  • Automated image analysis for monitoring vital fish habitats
  • EfficientNetV2 for marine echinoderm classification
  • Multi-mode CNNs for global chlorophyll-a concentration prediction
  • Spatio-temporal transformer
The ChatGPT summary is generated out of the summaries of the individual article abstracts.