AI Classification of Venomous Snake Images
The AI recognition of venomous snake images allows users to send pictures via LINE, and the system (LINE bot) returns the interpretation results. Currently, there are nearly 3,000 users, and it's also possible to consult venomous snake experts in real-time.
Design Concept: Combining LINE with AI Model
The theme MAP's design concept combines the immediacy of LINE with the accuracy of the AI model to achieve the goal of precision medicine.
LINE Function and Usage:
Besides the recognition of venomous snake images, the LINE bot can also perform web crawling to search for real-time antivenom data and relay it to users. With the user's consent, it can consult experts, providing medical information and care to patients bitten by venomous snakes. Currently, there are 3,000 LINE users.
Continuously Collecting Images to Improve the Model
The ultimate goal of AI venomous snake image recognition is to achieve wound classification. The length of teeth and toxins of different venomous snakes result in different wound appearances. In the future, we hope to 1. Use the pictures sent by users to further improve the model's accuracy. 2. Collaborate with various medical institutions to assist medical personnel in using it. 3. If there are enough wound photos, we can achieve a wound classification model, which will be useful even for patients who couldn't take a picture of the snake, enabling them to use the correct antivenom.