Identification and Instance Segmentation of Oil Spills Using DeepNeural Networks
Bibliographic record
Abstract
Oil spills impact natural and built environments, people and communities, food chain, and wildlife, and the road to full recovery is often long and costly to businesses, contractors, communities, and local governments. Existing oil spill detection methods including in-situ measurements and remote sensing primarily depend on involving skilled personnel in data collection, processing, and analysis, which could be expensive, slow, and subjective (influenced by prior experience, and judgment of problem parameters or solution space). In addition, oil pipelines and platforms can be located in remote and harsh areas, making it difficult and even hazardous for engineers to conduct timely inspections. Applying artificial intelligence (AI) can streamline this process and create more objective measures of oil spill and leakage detection. In this research, deep learning models, namely VGG-16 and mask R-CNN (mask regionbased convolutional neural network) are employed to identify and locate oil spills. These models represent state-of-the-art object recognition algorithms in computer vision. Red-green-blue (RGB) training images are collected using semi-supervised learning (i.e., keyword search) from the web. This initial visual dataset consists of a diverse set of photos taken by unmanned aerial vehicles (UAVs) or first-person cameras from previous oil spill accidents. The methodology consists of model training and validation, image classification, object detection, and semantic segmentation. The VGG16 model is used for image classification (to predict the existence of oil spill in an image) and yields an accuracy of ~93%. The mask R-CNN model is used for instance segmentation (to detect oil spills and marking their boundaries at pixel-level) and yields average precision and recall of 61% and 70%, respectively. Results can create opportunities for advancing the current practice of integrating AI and data analytics into downstream and upstream operations in the oil and gas industry, as well as enabling non-intrusive techniques for detection of environmental pollutants.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".