IMAGE CLASSIFICATION is a central problem in Computer Vision and Machine Learning.
Image Classification is the task of assigning a label to an image from a predefined set of categories. Given an input image, we return a label that categorizes the image (or the n most likely predicted labels for the object in the image).
Let's consider an input image, in this case the image of a dog.
The goal here is to take this input image and assign a label to it.
In our example, the classification algorithm is able to assign multiple labels to the image via probabilities.
At the end, we can take for instance the 3 most likely category labels for the object identified in our image as a result of the Image Classification task.
- Image Classification with pre-trained VGG16 (Notebook)
- image_class_1.jpg (input image of classification task)
- image_class_2.jpg (input image of classification task)
- image_class_3.jpg (input image of classification task)
- image_class_4.jpg (input image of classification task)
- image_class_5.jpg (input image of classification task)
- image_class_6.jpg (input image of classification task)
- image_class_7.jpg (input image of classification task)
- VGG16.png (image of VGG16 architecture)
- VGG16_2.png (image of VGG16 architecture)
- git_img_class.png (readme image)
- git_img_class_2.png (readme image)
Image Classification Model: VGG16
Libraries: Tensorflow, Keras, Matplotlib
Language: Python