The idea is to have a static site where people can visit. In this instance, the viewers are usually random people with the same interest as the original poster. People can then go to their web page and see what they posted. There are some people who post things for their satisfaction. However, not all social media apps are that way. You share these things with your friends, and usually provide context to what you are doing. Learning Multiple Layers of Features from Tiny Images, Alex Krizhevsky, 2009.Social media is all about sharing your ideas, music, pictures, and videos.Please cite it if you intend to use this dataset.
This tech report (Chapter 3) describes the dataset and the methodology followed when collecting it in much greater detail. The first 50000 lines correspond to the training set, and the last 10000 lines correspond "0" stands for an image that is not from the tiny db. Where the first image in the tiny db is indexed "1". The file has 60000 rows, each row contains a single index into the tiny db, Sivan Sabato was kind enough to provide this file, which maps CIFAR-100 images to images in the 80 million tiny images dataset. Indices into the original 80 million tiny images dataset The binary version of the CIFAR-100 is just like the binary version of the CIFAR-10, except that each image has two label bytes (coarse and fine) and 3072 pixel bytes, so the binary files look like this: The python and Matlab versions are identical in layout to the CIFAR-10, so I won't waste space describing them here. Yes, I know mushrooms aren't really fruit or vegetables and bears aren't really carnivores.ĬIFAR-100 binary version (suitable for C programs) Lawn-mower, rocket, streetcar, tank, tractor Orchids, poppies, roses, sunflowers, tulipsĪpples, mushrooms, oranges, pears, sweet peppersĬlock, computer keyboard, lamp, telephone, televisionīee, beetle, butterfly, caterpillar, cockroachĬamel, cattle, chimpanzee, elephant, kangarooĬrocodile, dinosaur, lizard, snake, turtleīicycle, bus, motorcycle, pickup truck, train Here is the list of classes in the CIFAR-100: SuperclassĪquarium fish, flatfish, ray, shark, trout Each image comes with a "fine" label (the class to which it belongs) and a "coarse" label (the superclass to which it belongs). The 100 classes in the CIFAR-100 are grouped into 20 superclasses. There are 500 training images and 100 testing images per class. This dataset is just like the CIFAR-10, except it has 100 classes containing 600 images each. The class name on row i corresponds to numeric label i. It is merely a list of the 10 class names, one per row.
This is an ASCII file that maps numeric labels in the range 0-9 to meaningful class names.
Therefore each file should be exactly 30730000 bytes long. The values are stored in row-major order, so the first 32 bytes are the red channel values of the first row of the image.Įach file contains 10000 such 3073-byte "rows" of images, although there is nothing delimiting the rows. The first 1024 bytes are the red channel values, the next 1024 the green, and the final 1024 the blue. The next 3072 bytes are the values of the pixels of the image. In other words, the first byte is the label of the first image, which is a number in the range 0-9.
0 kommentar(er)
