Revision history [back]

I have spent hours getting these pictures. They are Black & White and 100x100 pixels. These are good for training good detectors for abstract objects like a checkerboard, or maybe a specific logo. They are pretty random, I got them from the MISC. section of (https://en.wikipedia.org/wiki/ImageNet). The total number of images is 180,049. There are no duplicates (not just names but image content), and extremely few pictures that are error messages (1-5 maybe). I have spent multiple hours cleaning them up. I have also included "bg.txt" which says every picture's name and its location. I haven't seen anyone else provide something like this before online, so I thought I would share. I have used this and 1,000,000 positive pictures to make a good cascade object detector for detecting a 5x5 black and white checkerboard. If you are interested in that, I can put that in the link as well. Sorry It is not a question. Serves more to answer questions. You can point questions to this

http://linktrack.info/.2s775 (It tracks how many times it has been clicked. Links to Google Drive folder, take my word for it. I love statistics!)

I have spent hours getting these pictures. They are Black & White and 100x100 pixels. These are good for training good detectors for abstract objects like a checkerboard, or maybe a specific logo. They are pretty random, I got them from the MISC. section of (https://en.wikipedia.org/wiki/ImageNet). Image-net (www.image-net.org/). The total number of images is 180,049. There are no duplicates (not just names but image content), and extremely few pictures that are error messages (1-5 maybe). I have spent multiple hours cleaning them up. I have also included "bg.txt" which says every picture's name and its location. I haven't seen anyone else provide something like this before online, so I thought I would share. I have used this and 1,000,000 positive pictures to make a good cascade object detector for detecting a 5x5 black and white checkerboard. If you are interested in that, I can put that in the link as well. Sorry It is not a question. Serves more to answer questions. You can point questions to this

http://linktrack.info/.2s775 (It tracks how many times it has been clicked. Links to Google Drive folder, take my word for it. I love statistics!)

I have spent hours getting these pictures. They are Black & White and 100x100 pixels. These are good for training good detectors for abstract objects like a checkerboard, or maybe a specific logo. They are pretty random, I got them from the MISC. section of Image-net (www.image-net.org/). The total number of images is 180,049. There are no duplicates (not just names but image content), and extremely few pictures that are error messages (1-5 maybe). I have spent multiple hours cleaning them up. I have also included "bg.txt" which says every picture's name and its location. I haven't seen anyone else provide something like this before online, so I thought I would share. I have used this and 1,000,000 positive pictures to make a good cascade object detector for detecting a 5x5 black and white checkerboard. If you are interested in that, I can put that in the link as well. Sorry It is not a question. Serves more to answer questions. You can point questions to this

The pictures are in this link http://linktrack.info/.2s775 (It tracks how many times it has been clicked. Links to Google Drive folder, folder with the pictures compressed into a .zip file and bg.txt, take my word for it. I love statistics!)

I have spent hours getting these pictures. They are Black & White and 100x100 pixels. These are good for training good detectors for abstract objects like a checkerboard, or maybe a specific logo. They are pretty random, I got them from the MISC. section of Image-net (www.image-net.org/). The total number of images is 180,049. There are no duplicates (not just names but image content), and extremely few pictures that are error messages (1-5 maybe). I have spent multiple hours cleaning them up. I have also included "bg.txt" which says every picture's name and its location. I haven't seen anyone else provide something like this before online, so I thought I would share. I have used this and 1,000,000 positive pictures to make a good cascade object detector for detecting a 5x5 black and white checkerboard. If you are interested in that, I can put that in the link as well. Sorry It is not a question. Serves more to answer questions. You can point questions to this

The pictures are in this link http://linktrack.info/.2s775 (It tracks how many times it has been clicked. Links to Google Drive folder with the pictures compressed into a .zip file and bg.txt, take my word for it. I love statistics!)