Classifying Tattoos with Neural Networks

With nearly 1.5 billion monthly visitors and 346,000 pictures of tattoos, Tattoodo is taking advantage of deep learning to help categorize the growing number of uploaded images.

“At Tattoodo, we spend a lot of time and effort on classifying the tattoo pictures that are uploaded,” mentioned Goran Vuksic, a developer at Tattoodo. “A community member is able to provide a textual description and tag the tattoo with arbitrary hashtags, which obviously is a lot of responsibility to put in the hands of one member.”

To help tackle this problem, Vuksic and a few others came up with a deep learning project during the company’s “Hacker Day” in hopes to  classify the remainder of the images with missing descriptions and also to suggest hashtags during the upload and edit process.

Using DIGITS, GPUs on the Amazon cloud and the cuDNN-accelerated Caffe deep learning framework, the developers trained their neural networks on 13 unique styles of tattoos and their related common hashtags.

The internal hacker project is still in beta as they continue to improve results, but plan on rolling it into production once they achieve the highest possible accuracy.

Read more >

About Brad Nemire

Brad Nemire
Brad Nemire is on the Developer Marketing team and loves reading about all of the fascinating research being done by developers using NVIDIA GPUs. Reach out to Brad on Twitter @BradNemire and let him know how you’re using GPUs to accelerate your research. Brad graduated from San Diego State University and currently resides in San Jose, CA. Follow @BradNemire on Twitter