Blog

Edited on August 30, 2019. After we had a functional transfer learning model, we were feeling good about Parasite ID. Then, somehow, Cameron accidentally uploaded a picture of a hammock that Vicki had originally taken on her vacation to the web app, and the app predicted that the hammock had schistosomiasis. This was a very... Read More »

Edited on August 24, 2019. Our classification model is ambitious; we felt that it was important to try to classify all of the eggs of the nine most common parasitic worms which cause neglected tropical disease. This has resulted in a 10-class classification model (nine egg-types plus the negative class). Currently, our dataset is composed... Read More »

      The WHO estimates that over a quarter of the world’s population is infected by soil-transmitted helminths1. By the numbers, schistosomiasis is the deadliest of these diseases. Symptoms include2: Diarrhea Bloody stools Enlargement of abdomen Splenomegaly Weakness Anemia Inflamed liver       The Global Burden of Disease Study estimated more than 200,000 annual deaths related to schistosomiasis3, with... Read More »

      Though AWS Lambda was an elegant solution, it has strict constraints on the size of the code and memory used. Because the Keras and TensorFlow libraries require so much space on their own, there is not a lot of space left for a model. As we continued to train new and better models, they grew... Read More »

      AWS Lambda was a clear choice for our model delivery system, due to its serverless architecture and low cost. However, Lambda functions can only be 250 MB unzipped, which posed a problem to us. Parasite ID’s first classification algorithm was a neural network with almost 18 million parameters! It requires Keras (built on TensorFlow) to... Read More »