Classification of Dog Personalities

As the owner of three unique children, I know that there are many differences between dogs. Everyone who owns a dog has his own idea, so I know that there is personality. I have kept a lot of dogs in my life, and I found that dogs can be divided into three categories. Anne is a princess of the dog world, Anne dog has a loving, manipulated character. Annie dog plays a role of the queen bee. As a leader in the backpack, she is consistent with everyone. She dominates all other dogs, but usually due to physical grazing or malicious roar.

Returning to the example of dog classification, you can think that a child erroneously treats a cat as a dog. It saw a cat and said it was a dog based on its classification: four feet, a furry-like body, and a tail. The child is fixed by his parents and tells the animal that it is actually a cat. Next time a child sees a cat, he never calls it a dog. In other words, we need to collect as many different receipts as possible to train our computer to actually read the receipt. If the algorithm works well enough, we have a "learning" artificial neural network that can handle new inputs (new receipts). This artificial neural network is said to have learned from several examples (receipt with mark) and error (error propagation).

Part 1: The robot is reading a receipt! You can learn all other knowledge about deep learning on the Sensibill blog at WordPress.com.

Standard classification is used in almost all classification models. The input is entered into a series of layers, and the class probability is output at the end. If you want to predict a cat's dog, you can train the model with a similar (but not identical) dog / cat picture that is expected within the predicted time. Of course, this requires a data set similar to what is expected when using the model for prediction. A good example is face recognition. For a small number of people, we train one shot classification model with dataset including various angles, lighting etc Next, if you want to check whether the person X appears in the image, take a picture of that person and ask the model if the person is in the image (model does not use someone's photo for training X Hmm).

It classifies CNN and shoots an image and identifies what is contained in the image. For example, if you have several pictures of cats and dogs, you can train the classification of CNN and decide which is which. In the classifier CNN we are using, the task is to classify small images into one of 1000 categories, which is a very popular task called ImageNet. The classifier network performs multiple convolutions on the image to generate useful classification features. As the image passes through the network, its size (pixels) gets smaller and smaller but the structure of each pixel gets bigger. Beginning with full-resolution RGB input (3 elements), it iteratively reduces the image to a single pixel, but there are many elements: the image represents the probability of each category