All datasets mentioned here are free for research and educational purposes.
| Description | DS Size | Formats |
|---|---|---|
| Research Google | 2.1M | .csv, YouTube urls |
| Kaggle | 5k+ | - |
| AWS public Datasets | k+ | - |
| UCI | k+ | - |
| Description | DS Size | Formats |
|---|---|---|
| Project Common Voice | 400k+ | .csv,.mp3 |
| VoxForge | 10k+ | - |
| The CMU Audio Databases | 1k | .raw |
| Description | DS Size | Formats |
|---|---|---|
| Flickr Data Yahoo dataset | 100M img, 1M videos | - |
| Tiny Images | 80M | tiny img |
| CIFAR-10 and CIFAR-100 | 80M | 3kB img |
| IMAGENET | 14M | img |
| VQA Visual Question Answering | 10M+ | - |
| Open Images Dataset | 9M | img urls |
| YouTube-8M Dataset | 8M | YouTube urls |
| Street View House Numbers | 600k | .mat |
| Microsoft COCO | 330k | .txt, img |
| COCO-QA Image Question answering | 123k | .txt, .jpg |
| CAVIAR video sequences of mall and public space behavior | 90k+ | .mpeg, .jpg |
| MNIST Handwritten digits | 60k | .msb |
| Flickr 30k | 30k img, 150k captions | - |
| OULU dataset of eyegazing, face expressions | ~20k | - |
| Berkeley Segmentation DS | 12k | .cdr |
| Traffic Image Sequences | 8k | grayscale .gif |
| Sign Language Recognition | 5k | img |
| AR Face Database | 4k | .raw, .avi |
| Fingerprint database | <1k | grayscale .gif |
| Description | DS Size | Formats |
|---|---|---|
| Question Answering Corpus | 1M+ Q&A | .txt, .html |
| Fictional Conversations Extracted from Raw Movie Scripts | 220k conversations | .txt |
| Microsoft's Social Media Conversation Corpus | 12k tweets | .txt |
| Description | DS Size | Formats |
|---|---|---|
| University of South Florida Digital Mammography | 2620 cases | .ics |
| Bone and joint CT scan images | - | .dcm |