Canine/Baked-Good Identification
Perhaps one of the most horrific things to come out of 2016 is a query by one Karen Zack over Twitter. Tapping into a primal confusion once thought unthinkable, the Internet was plunged into furious debate over whether her series of pictures depicted a Chihuahua, or a most handsome muffin.
As the Uber of canine/baked-good identification, we have come to upend the meme sharing economy and deliver unseen profits to our shareholders. We’re a cutting-edge company at the crest of the ML revolution, and we know what it’s like to create something truly historic.
Our proprietary algorithm, developed in house by industry-leading research scientists, delivers a vital service to the human race. Utilizing the Caffe ML library and a massive doggo database, we have trained our gradient descent algorithms to be both efficient and versatile. What was predicted long ago by The Great Messiah™ has now become reality by harnessing the awesome power of convolutional neural networks, small big data, and deep learning.
We were able to achieve an incredible classification accuracy of 66%: Using 17 images of canines, 32 images of muffins, 8000 Iterations, and an 8 layer neural net.
Looking ahead, we look forward to solving more complex image catagorization problems such such as puppy or bagel paradox, and the parrot or guacamole connundrum.
