Awesome Deep Learning
by ChristosChristofidis · ChristosChristofidis/awesome-deep-learning
A curated list of awesome Deep Learning tutorials, projects and communities.
by Yoshua Bengio, Ian Goodfellow and Aaron Courville (05/07/2015)
by François Chollet with Tomasz Kalinowski and J. J. Allaire
by Aurélien Géron Oct 15, 2019
N
MMachine Learning and Deep Learning Courses from Amazon's Machine Learning university
by Fei-Fei Li, Andrej Karpathy (2017)
A free five-weekend plan to self-learners to learn the basics of deep-learning architectures like CNNs, LSTMs, RNNs, VAEs, GANs, DQN, A3C and more (2019)
List of Deep Learning online courses (some are free) from Classpert Online Course Search
Breaking into AI with the best course from Andrew NG.
by Geoffrey Hinton, Yoshua Bengio, Yann LeCun, Andrew Ng, Nando de Freitas and several others @ IPAM, UCLA (2012)
A great introductory course on Deep Learning by Udacity and Facebook AI
A seven day bootcamp designed in MIT to introduce deep learning methods and applications (2019)
A free deep reinforcement learning course by OpenAI (2019)
MSc in Artificial Intelligence for the University of Amsterdam.
by Shimon Ullman, Tomaso Poggio, Ethan Meyers @ MIT (2013)
The Air Freight data set is a ray-traced image sequence along with ground truth segmentation based on textural characteristics. (455 images + GT, each 160x120 pixels). (Formats: PNG)
ALOI is a color image collection of one-thousand small objects, recorded for scientific purposes. In order to capture the sensory variation in object recordings, we systematically varied viewing angle, illumination angle, and illumination color for each object, and additionally captured wide-baseline stereo images. We recorded over a hundred images of each object, yielding a total of 110,250 images for the collection. (Formats: png)
Most images & annotations are supplemented by various ASM/AAM analyses using the AAM-API. (Formats: bmp,asf)
Contains 450K affective annotations of emotional responses and linguistic explanations for 80,000 artworks of WikiArt.
A variety of datasets including geons, objects, and "greebles". Good for testing recognition algorithms. (Formats: pict)
about 20 images - mostly top-down views of small objects and toys. (Formats: GIF)
90K video frames in 90 sequences of various human activities, with XML ground truth of detection and behavior classification (Formats: MPEG2 & JPEG)
A database of 41,368 face images of 68 people captured under 13 poses, 43 illuminations conditions, and with 4 different expressions.
Images, sequences, stereo pairs (thousands of images) (Formats: Sun Rasterimage)
Texture and reflectance measurements for over 60 samples of 3D texture, observed with over 200 different combinations of viewing and illumination directions. (Formats: bmp)
A dataset oriented towards computational color constancy, but useful for computer vision in general. It includes synthetic data, camera sensor data, and over 700 images. (Formats: tiff)
11 sets of color images for testing algorithms for content-based retrieval. Most sets have a description file with names of objects in each image. (Formats: jpg)
DTextual QA corpus from CNN and DailyMail. More than 300K documents in total. Paper for reference.
Densely sampled view spheres - upper half of the view sphere of two toy objects with 2500 images each. (Formats: tiff)
Digital embryos are novel objects which may be used to develop and test object recognition systems. They have an organic appearance. (Formats: various formats are available on request)
Images and Videos of his-res of studies taken from Gastrointestinal Video endoscopy. (Formats: jpg, mpg, gif)
FContains about 10 million news articles classified using opensources.co types
FMNIST like fashion product dataset consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes.
Database contains 1002 face images showing subjects at different ages. (Formats: jpg)
FVC2000 is the First International Competition for Fingerprint Verification Algorithms. Four fingerprint databases constitute the FVC2000 benchmark (3520 fingerprints in all).
The database contains 35 gestures and consists of 1400 image sequences that contain gestures of 20 different persons recorded under non-uniform daylight lighting conditions. (Formats: mpg,jpg)
4000+ 1536x1024 (16 bit) calibrated outdoor images (Formats: homebrew)
2 turntable sequences from different viewing heights, 36 images each, resolution 1000x750, color (Formats: PPM)
Images obtained from a variety of imaging modalities -- raw CFA images, range images and a host of "medical images". (Formats: homebrew)
The JAFFE database consists of 213 images of Japanese female subjects posing 6 basic facial expressions as well as a neutral pose. Ratings on emotion adjectives are also available, free of charge, for research purposes. (Formats: TIFF Grayscale images.)
Contains over 800,000 diverse fashion images. Each image in this dataset is labeled with 50 categories, 1,000 descriptive attributes, bounding box and clothing landmarks
L15488 visible-infrared paired images (30976 images) for low-light vision research, ProjectPage
Images from the textbook by Jain, Kasturi, Schunck (20+ images) (Formats: GIF TIFF)
100 or more images of mammograms with ground truth. Additional images available by request, and links to several other mammography databases are provided. (Formats: homebrew)
Six multi-frame stereo data sets of scenes containing planar regions. Each data set contains 9 color images and subpixel-accuracy ground-truth data. (Formats: ppm)
High Altitude Imagery from around the world for environmental modeling in support of NASA EOS program (Formats: JPG and HDF)
MOver over 5 million images from 5 different domains for multi-source ocr/text recognition DA research, ProjectPage
Over 55,000 3D CAD and solid models of (mostly) mechanical/machined engineering designs. (Formats: gif,vrml,wrl,stp,sat)
Color, CAT and MRI image samples - over 30 images (Formats: jpeg)
OOpen Images is a dataset of ~9 million URLs to images that have been annotated with labels spanning over 6000 categories.
several sets of 3D object models collected over several years to use in object recognition research (Formats: homebrew, vrml)
Hundreds of real and synthetic images (Formats: gif, homebrew)
Over 1000 range images, 3D object models, still images and motion sequences (Formats: gif, ppm, vrml, homebrew)
Synthetic and real sequences with machine-readable ground truth optical flow fields, plus tools to generate ground truth for new sequences. (Formats: ppm,tif,homebrew)
This is the first 3D texture database which provides both full real surface rotations and registered photometric stereo data (30 textures, 1680 images). (Formats: TIFF)
Two large-scale and complementary visio-linguistic datasets (aka Nr3D and Sr3D) for identifying fine-grained 3D objects in ScanNet scenes. Nr3D contains 41.5K natural, free-form utterances, and Sr3d contains 83.5K template-based utterances.
SANAD Dataset is a large collection of Arabic news articles that can be used in different Arabic NLP tasks such as Text Classification and Word Embedding. The articles were collected using Python scripts written specifically for three popular news websites: AlKhaleej, AlArabiya and Akhbarona.
Ssynthetic sequence for testing structure from motion algorithms (Formats: pgm)
9 synthetic sequences designed for testing motion analysis applications, including full ground truth of motion and camera parameters. (Formats: gif)
Stanford released ~100,000 English QA pairs and ~50,000 unanswerable questions
a small set of synthetic images of a hallway with varying amounts of noise added. Use these images to benchmark your stereo algorithm. (Formats: raw, viff (khoros), or tiff)
A collection of synthetic range images taken from high-resolution polygonal models available on the web (Formats: homebrew)
Contains over 4,000 color images corresponding to 126 people's faces (70 men and 56 women). Frontal views with variations in facial expressions, illumination, and occlusions. (Formats: RAW (RGB 24-bit))
Database for testing multiclass object detection and scene recognition algorithms. Over 72,000 images with 2873 annotated frames. More than 50 annotated object classes. (Formats: jpg)
A collection of over 300 real images of 100 objects taken under three different illuminaiton conditions (Diffuse/Ambient/Directed). -- Use these images to test algorithms for detecting and compensating specular highlights in color images. (Formats: TIFF )
The XM2VTSDB contains four digital recordings of 295 people taken over a period of four months. This database contains both image and video data of faces.
thousands of frames of digitized traffic image sequences as well as the 'Marbled Block' sequence (grayscale images) (Formats: GIF)
Includes classifications - 1000+ color images (Formats: ppm)
a benchmark database for image retrieval with predefined ground truth. (Formats: tiff)
Large image database with aerial, space, stereo, medical images and more. (Formats: homebrew)
contains color images of faces under different illuminants and camera calibration conditions as well as skin spectral reflectance measurements of each person.
Database of 320 surface textures, each captured under three illuminants, six spatial resolutions and nine rotation angles. A set of test suites is also provided so that texture segmentation, classification, and retrieval algorithms can be tested in a standard manner. (Formats: bmp, ras, xv)
80 image sets (Formats: Sun rasterimage)
Images of 8 objects seen from many different view points. The view sphere is sampled using a geodesic with 172 images/sphere. Two sets for training and testing are available. (Formats: ppm)
VOC2012 dataset containing 12k images with 20 annotated classes for object detection and segmentation.
Thousands of images of a cart, ladder, stool, bicycle, chairs, and cluttered scenes with ground truth labelings of edges and regions. (Formats: jpg)
165 images (15 individuals) with different lighting, expression, and occlusion configurations.
5760 single light source images of 10 subjects each seen under 576 viewing conditions (9 poses x 64 illumination conditions). (Formats: PGM)
YouTube-8M is a large-scale labeled video dataset that consists of 8 million YouTube video IDs and associated labels from a diverse vocabulary of 4800 visual entities.
A
B
B
C
C↗
C
C↗
D
D
D
D
D
D
F
F
H
H
I↗
J
K
L
L
M
M
MApplication-oriented deep reinforcement learning framework addressing real-world decision problems.
M
M
M
N
N
N
O
P
P
Q
S
S
S
T
T
V
A
A
A
A↗
ACurated list of articles related to deep learning scientific research applied to music
A
ACurated list of articles related to deep learning scientific research on graph structured data at the graph level.
ACurated list of articles related to deep learning scientific research on graph structured data at the node level.
B
C
DEasy to install and use deep learning Faster R-CNN face detection for images and video in a docker container.
F
G
L
M
M
Mcontains examples, utilities and best practices for building recommendation systems. Implementations of several state-of-the-art algorithms are provided for self-study and customization in your own applications.
P
R↗
TAndrej Karpathy blog post about using RNN for generating text.
T
T
T
W
A
GBrowser extension (Chrome and Firefox) that automatically finds and links to code implementations for ML papers anywhere online: Google, Twitter, Arxiv, Scholar, etc.
Community platform for Open Source ML – Manage experiments, data & models and create collaborative ML projects easily.
DDeep learning training platform with integrated support for distributed training, hyperparameter tuning, smart GPU scheduling, experiment tracking, and a model registry.
DA little logger for machine learning research. Log any object to the console, CSVs, TensorBoard, text log files, and more with just one call to logger.log()
DVC is built to make ML models shareable and reproducible. It is designed to handle large files, data sets, machine learning models, and metrics as well as code.
HFastest unstructured dataset management for TensorFlow/PyTorch by activeloop.ai. Stream & version-control data. Converts large data into single numpy-like array on the cloud, accessible on any machine.
MMLEM is a tool to easily package, deploy and serve Machine Learning models. It seamlessly supports a variety of scenarios like real-time serving and batch processing.
NEasy-to-use library to boost deep learning inference leveraging multiple deep learning compilers.
N
T
C
C
H
K
M
P
T
T
T
T
ABy Natalie Hammel and Lorraine Yurshansky
a series of mini-lectures by Leo Isikdogan on YouTube (2018)
a live video course that teaches how to apply deep learning to text and images using the powerful Keras library and its R language interface.
by Steve Jurvetson (and panel) at VLAB in Stanford.
This tutorial is styled as a graduate lecture about medical imaging with deep learning. This will cover the background of popular medical image domains (chest X-ray and histology) as well as methods to tackle multi-modality/view, segmentation, and counting tasks.