Entity identification using machine learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to accurately identify and track individual fish, particularly within large populations, due to similarities in appearance and the difficulty in distinguishing between fish of the same species, which hinders monitoring and management in aquaculture settings.
Innovation Solution
A machine learning model is trained to recognize fish using their unique characteristics, such as spot patterns, and generates embeddings that are mapped to a high-dimensional space for identification and re-identification, allowing for the tracking of fish over time and the detection of health issues or diseases within fish populations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional visual inspection methods are used to identify fish, then the process is simple and requires minimal technology, but the ability to distinguish between individual fish of the same species is insufficient
Solution Approach 1:
The patent replaces manual visual inspection with an automated machine learning system that uses image processing and neural networks to identify fish. The system captures images of fish and uses a trained machine learning model to generate unique identifiers based on visual characteristics, eliminating the need for human experts to manually distinguish between individual fish.
Solution Approach 2:
The system creates digital copies (embeddings) of fish visual characteristics that can be stored and compared. Instead of physically examining fish, the system generates vector representations of fish images and compares these digital copies to identify and track individual fish across multiple observations.
2Productivity
If manual tracking of individual fish is attempted in large populations, then direct observation is possible, but the scale and complexity of tracking becomes unmanageable
Solution Approach 1:
The machine learning system automatically performs the entire tracking process without human intervention. It captures images, processes them through the neural network, generates identifiers, and maintains tracking records autonomously, allowing the system to scale to large fish populations without requiring proportional increases in human labor.
Solution Approach 2:
The system transforms the tracking problem from manual observation to automated image parameter analysis. By converting visual characteristics into numerical embeddings and using distance metrics in vector space, the system can efficiently compare and track fish at scale without the complexity of manual population management.
3Measurement precision
If high-resolution images are generated to improve fish identification, then identification accuracy increases, but processing time and computational resources increase
Solution Approach 1:
The system performs preliminary actions by pre-processing images to generate high-resolution representations before identification. It creates enhanced image versions and generates embeddings in advance, so that when identification is needed, the computationally intensive work has already been completed, reducing real-time processing delays.
Solution Approach 2:
The system transitions from working directly with pixel data to operating in an embedding vector space. By projecting high-resolution image features into a lower-dimensional vector representation, the system maintains identification accuracy while reducing the computational burden of comparing and processing fish images.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media for identification and re-identification of fish. In some implementations, first media representative of aquatic cargo is received. Second media based on the first media is generated, wherein a resolution of the second media is higher than a resolution of the first media. A cropped representation of the second media is generated. The cropped representation is provided to the machine learning model. In response to providing the cropped representation to the machine learning model, an embedding representing the cropped representation is generated using the machine learning model. The embedding is mapped to a high dimensional space. Data identifying the aquatic cargo is provided to a database, wherein the data identifying the aquatic cargo comprises an identifier of the aquatic cargo, the embedding, and a mapped region of the high dimensional space.