Deep Learning for Accurate Vision-based Catch Composition in Tropical Tuna Purse Seiners

Purse seiners play a crucial role in tuna fishing, as approximately 69% of the world's tropical tuna is caught using this gear. All tuna Regional Fisheries Management Organizations have established minimum standards to use electronic monitoring (EM) in fisheries in addition to traditional observers. The EM systems produce a massive amount of video data that human analysts must process. Integrating artificial intelligence (AI) into their workflow can decrease that workload and improve the accuracy of the reports. However, species identification still poses significant challenges for AI, as achieving balanced performance across all species requires appropriate training data. Here, we quantify the difficulty experts face to distinguish bigeye tuna (BET, Thunnus Obesus) from yellowfin tuna (YFT, Thunnus Albacares) using images captured by EM systems. We found inter-expert agreements of 42.9% $\pm$ 35.6% for BET and 57.1% $\pm$ 35.6% for YFT. We then present a multi-stage pipeline to estimate the species composition of the catches using a reliable ground-truth dataset based on identifications made by observers on board. Three segmentation approaches are compared: Mask R-CNN, a combination of DINOv2 with SAM2, and a integration of YOLOv9 with SAM2. We found that the latest performs the best, with a validation mean average precision of 0.66 $\pm$ 0.03 and a recall of 0.88 $\pm$ 0.03. Segmented individuals are tracked using ByteTrack. For classification, we evaluate a standard multiclass classification model and a hierarchical approach, finding a superior generalization by the hierarchical. All our models were cross-validated during training and tested on fishing operations with fully known catch composition. Combining YOLOv9-SAM2 with the hierarchical classification produced the best estimations, with 84.8% of the individuals being segmented and classified with a mean average error of 4.5%.

Review Mining for Feature Based Opinion Summarization and Visualization

The application and usage of opinion mining, especially for business intelligence, product recommendation, targeted marketing etc. have fascinated many research attentions around the globe. Various research efforts attempted to mine opinions from customer reviews at different levels of granularity, including word-, sentence-, and document-level. However, development of a fully automatic opinion mining and sentiment analysis system is still elusive. Though the development of opinion mining and sentiment analysis systems are getting momentum, most of them attempt to perform document-level sentiment analysis, classifying a review document as positive, negative, or neutral. Such document-level opinion mining approaches fail to provide insight about users sentiment on individual features of a product or service. Therefore, it seems to be a great help for both customers and manufacturers, if the reviews could be processed at a finer-grained level and presented in a summarized form through some visual means, highlighting individual features of a product and users sentiment expressed over them. In this paper, the design of a unified opinion mining and sentiment analysis framework is presented at the intersection of both machine learning and natural language processing approaches. Also, design of a novel feature-level review summarization scheme is proposed to visualize mined features, opinions and their polarity values in a comprehendible way.

Subjectivity Classification using Machine Learning Techniques for Mining Feature-Opinion Pairs from Web Opinion Sources

Due to flourish of the Web 2.0, web opinion sources are rapidly emerging containing precious information useful for both customers and manufactures. Recently, feature based opinion mining techniques are gaining momentum in which customer reviews are processed automatically for mining product features and user opinions expressed over them. However, customer reviews may contain both opinionated and factual sentences. Distillations of factual contents improve mining performance by preventing noisy and irrelevant extraction. In this paper, combination of both supervised machine learning and rule-based approaches are proposed for mining feasible feature-opinion pairs from subjective review sentences. In the first phase of the proposed approach, a supervised machine learning technique is applied for classifying subjective and objective sentences from customer reviews. In the next phase, a rule based method is implemented which applies linguistic and semantic analysis of texts to mine feasible feature-opinion pairs from subjective sentences retained after the first phase. The effectiveness of the proposed methods is established through experimentation over customer reviews on different electronic products.