Zero-Shot Learning

The ImageNet dataset contains 14,197,122 annotated images according to the WordNet hierarchy. Since 2010 the dataset is used in the ImageNet Large Scale Visual Recognition Challenge (ILSVRC), a benchmark in image classification and object detection. The publicly released dataset contains a set of manually annotated training images. A set of test images is also released, with the manual annotations withheld. ILSVRC annotations fall into one of two categories: (1) image-level annotation of a binary label for the presence or absence of an object class in the image, e.g., “there are cars in this image” but “there are no tigers,” and (2) object-level annotation of a tight bounding box and class label around an object instance in the image, e.g., “there is a screwdriver centered at position (20,25) with width of 50 pixels and height of 30 pixels”. The ImageNet project does not own the copyright of the images, therefore only thumbnails and URLs of images are provided. * Total number of non-empty WordNet synsets: 21841 * Total number of images: 14197122 * Number of images with bounding box annotations: 1,034,908 * Number of synsets with SIFT features: 1000 * Number of images with SIFT features: 1.2 million Source: ImageNet Large Scale Visual Recognition Challenge Image Source: https://cs.stanford.edu/people/karpathy/cnnembed/

The CIFAR-100 dataset (Canadian Institute for Advanced Research, 100 classes) is a subset of the Tiny Images dataset and consists of 60000 32x32 color images. The 100 classes in the CIFAR-100 are grouped into 20 superclasses. There are 600 images per class. Each image comes with a "fine" label (the class to which it belongs) and a "coarse" label (the superclass to which it belongs). There are 500 training images and 100 testing images per class. The criteria for deciding whether an image belongs to a class were as follows: * The class name should be high on the list of likely answers to the question “What is in this picture?” * The image should be photo-realistic. Labelers were instructed to reject line drawings. * The image should contain only one prominent instance of the object to which the class refers. * The object may be partially occluded or seen from an unusual viewpoint as long as its identity is still clear to the labeler. Source: https://www.cs.toronto.edu/~kriz/cifar.html Image Source: https://www.cs.toronto.edu/~kriz/cifar.html

The Caltech-UCSD Birds-200-2011 (CUB-200-2011) dataset is the most widely-used dataset for fine-grained visual categorization task. It contains 11,788 images of 200 subcategories belonging to birds, 5,994 for training and 5,794 for testing. Each image has detailed annotations: 1 subcategory label, 15 part locations, 312 binary attributes and 1 bounding box. The textual information comes from Reed et al.. They expand the CUB-200-2011 dataset by collecting fine-grained natural language descriptions. Ten single-sentence descriptions are collected for each image. The natural language descriptions are collected through the Amazon Mechanical Turk (AMT) platform, and are required at least 10 words, without any information of subcategories and actions. Source: Fine-grained Visual-textual Representation Learning Image Source: http://www.vision.caltech.edu/visipedia/CUB-200-2011.html

UCF101 dataset is an extension of UCF50 and consists of 13,320 video clips, which are classified into 101 categories. These 101 categories can be classified into 5 types (Body motion, Human-human interactions, Human-object interactions, Playing musical instruments and Sports). The total length of these video clips is over 27 hours. All the videos are collected from YouTube and have a fixed frame rate of 25 FPS with the resolution of 320 × 240. Source: Two-stream Collaborative Learning with Spatial-Temporal Attention for Video Classification Image Source: https://www.crcv.ucf.edu/data/UCF101.php

ScanNet is an instance-level indoor RGB-D dataset that includes both 2D and 3D data. It is a collection of labeled voxels rather than points or objects. Up to now, ScanNet v2, the newest version of ScanNet, has collected 1513 annotated scans with an approximate 90% surface coverage. In the semantic segmentation task, this dataset is marked in 20 classes of annotated 3D voxelized objects. Source: A Review of Point Cloud Semantic Segmentation Image Source: http://www.scan-net.org/

The Kinetics dataset is a large-scale, high-quality dataset for human action recognition in videos. The dataset consists of around 500,000 video clips covering 600 human action classes with at least 600 video clips for each action class. Each video clip lasts around 10 seconds and is labeled with a single action class. The videos are collected from YouTube. Source: Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey

Oxford 102 Flower is an image classification dataset consisting of 102 flower categories. The flowers chosen to be flower commonly occurring in the United Kingdom. Each class consists of between 40 and 258 images. The images have large scale, pose and light variations. In addition, there are categories that have large variations within the category and several very similar categories.

The Describable Textures Dataset (DTD) contains 5640 texture images in the wild. They are annotated with human-centric attributes inspired by the perceptual properties of textures. Source: Where is the Fake? Patch-Wise Supervised GANs for Texture Inpainting Image Source: https://www.robots.ox.ac.uk/~vgg/data/dtd/

The HMDB51 dataset is a large collection of realistic videos from various sources, including movies and web videos. The dataset is composed of 6,766 video clips from 51 action categories (such as “jump”, “kiss” and “laugh”), with each category containing at least 101 clips. The original evaluation scheme uses three different training/testing splits. In each split, each action class has 70 clips for training and 30 clips for testing. The average accuracy over these three splits is used to measure the final performance. Source: Action Recognition with Trajectory-Pooled Deep-Convolutional Descriptors Image Source: https://serre-lab.clps.brown.edu/resource/hmdb-a-large-human-motion-database

The ActivityNet dataset contains 200 different types of activities and a total of 849 hours of videos collected from YouTube. ActivityNet is the largest benchmark for temporal activity detection to date in terms of both the number of activity categories and number of videos, making the task particularly challenging. Version 1.3 of the dataset contains 19994 untrimmed videos in total and is divided into three disjoint subsets, training, validation, and testing by a ratio of 2:1:1. On average, each activity category has 137 untrimmed videos. Each video on average has 1.41 activities which are annotated with temporal boundaries. The ground-truth annotations of test videos are not public. Source: Dynamic Temporal Pyramid Network: A Closer Look at Multi-Scale Modeling for Activity Detection

The Food-101 dataset consists of 101 food categories with 750 training and 250 test images per category, making a total of 101k images. The labels for the test images have been manually cleaned, while the training set contains some noise. Source: Combining Weakly and Webly Supervised Learning for Classifying Food Images Image Source: https://data.vision.ee.ethz.ch/cvl/datasets_extra/food-101/

The Stanford Cars dataset consists of 196 classes of cars with a total of 16,185 images, taken from the rear. The data is divided into almost a 50-50 train/test split with 8,144 training images and 8,041 testing images. Categories are typically at the level of Make, Model, Year. The images are 360×240. Source: View Independent Vehicle Make, Model and Color Recognition Using Convolutional Neural Network Image Source: https://ai.stanford.edu/~jkrause/cars/car_dataset.html

The dataset contains 400 human action classes, with at least 400 video clips for each action. Each clip lasts around 10s and is taken from a different YouTube video. The actions are human focussed and cover a broad range of classes including human-object interactions such as playing instruments, as well as human-human interactions such as shaking hands. Source: https://arxiv.org/abs/1705.06950 Image source: https://arxiv.org/abs/1705.06950

The Caltech101 dataset contains images from 101 object categories (e.g., “helicopter”, “elephant” and “chair” etc.) and a background category that contains the images not from the 101 object categories. For each object category, there are about 40 to 800 images, while most classes have about 50 images. The resolution of the image is roughly about 300×200 pixels. Source: Simple and Efficient Learning using Privileged Information

Eurosat is a dataset and deep learning benchmark for land use and land cover classification. The dataset is based on Sentinel-2 satellite images covering 13 spectral bands and consisting out of 10 classes with in total 27,000 labeled and geo-referenced images. Source: EuroSAT Image Source: EuroSAT

SemanticKITTI is a large-scale outdoor-scene dataset for point cloud semantic segmentation. It is derived from the KITTI Vision Odometry Benchmark which it extends with dense point-wise annotations for the complete 360 field-of-view of the employed automotive LiDAR. The dataset consists of 22 sequences. Overall, the dataset provides 23201 point clouds for training and 20351 for testing. Source: Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic Segmentation Image Source: https://github.com/PRBonn/semantic-kitti-api

FGVC-Aircraft contains 10,200 images of aircraft, with 100 images for each of 102 different aircraft model variants, most of which are airplanes. The (main) aircraft in each image is annotated with a tight bounding box and a hierarchical airplane model label. Aircraft models are organized in a four-levels hierarchy. The four levels, from finer to coarser, are: * Model, e.g. Boeing 737-76J. Since certain models are nearly visually indistinguishable, this level is not used in the evaluation. * Variant, e.g. Boeing 737-700. A variant collapses all the models that are visually indistinguishable into one class. The dataset comprises 102 different variants. * Family, e.g. Boeing 737. The dataset comprises 70 different families. * Manufacturer, e.g. Boeing. The dataset comprises 41 different manufacturers. The data is divided into three equally-sized training, validation and test subsets. Source: https://www.robots.ox.ac.uk/~vgg/data/fgvc-aircraft/ Image Source: Fine-Grained Visual Classification of Aircraft

The MPII Human Pose Dataset for single person pose estimation is composed of about 25K images of which 15K are training samples, 3K are validation samples and 7K are testing samples (which labels are withheld by the authors). The images are taken from YouTube videos covering 410 different human activities and the poses are manually annotated with up to 16 body joints. Source: 2D/3D Pose Estimation and Action Recognition using Multitask Deep Learning Image Source: http://human-pose.mpi-inf.mpg.de/

The Stanford 3D Indoor Scene Dataset (S3DIS) dataset contains 6 large-scale indoor areas with 271 rooms. Each point in the scene point cloud is annotated with one of the 13 semantic categories. Source: Grid-GCN for Fast and Scalable Point Cloud Learning Image Source: https://www.researchgate.net/figure/Examples-of-classified-scenes-in-S3DIS-dataset-left-with-groundtruth-right_fig2_328307943

NTU RGB+D is a large-scale dataset for RGB-D human action recognition. It involves 56,880 samples of 60 action classes collected from 40 subjects. The actions can be generally divided into three categories: 40 daily actions (e.g., drinking, eating, reading), nine health-related actions (e.g., sneezing, staggering, falling down), and 11 mutual actions (e.g., punching, kicking, hugging). These actions take place under 17 different scene conditions corresponding to 17 video sequences (i.e., S001–S017). The actions were captured using three cameras with different horizontal imaging viewpoints, namely, −45∘,0∘, and +45∘. Multi-modality information is provided for action characterization, including depth maps, 3D skeleton joint position, RGB frames, and infrared sequences. The performance evaluation is performed by a cross-subject test that split the 40 subjects into training and test groups, and by a cross-view test that employed one camera (+45∘) for testing, and the other two cameras for training. Source: Action Recognition for Depth Video using Multi-view Dynamic Images

The Charades dataset is composed of 9,848 videos of daily indoors activities with an average length of 30 seconds, involving interactions with 46 objects classes in 15 types of indoor scenes and containing a vocabulary of 30 verbs leading to 157 action classes. Each video in this dataset is annotated by multiple free-text descriptions, action labels, action intervals and classes of interacting objects. 267 different users were presented with a sentence, which includes objects and actions from a fixed vocabulary, and they recorded a video acting out the sentence. In total, the dataset contains 66,500 temporal annotations for 157 action classes, 41,104 labels for 46 object classes, and 27,847 textual descriptions of the videos. In the standard split there are7,986 training video and 1,863 validation video. Source: Temporal Reasoning Graph for Activity Recognition

The NUS-WIDE dataset contains 269,648 images with a total of 5,018 tags collected from Flickr. These images are manually annotated with 81 concepts, including objects and scenes. Source: Parallel Grid Pooling for Data Augmentation Image Source: Li et al

The PASCAL Context dataset is an extension of the PASCAL VOC 2010 detection challenge, and it contains pixel-wise labels for all training images. It contains more than 400 classes (including the original 20 classes plus backgrounds from PASCAL VOC segmentation), divided into three categories (objects, stuff, and hybrids). Many of the object categories of this dataset are too sparse and; therefore, a subset of 59 frequent classes are usually selected for use. Source: Image Segmentation Using Deep Learning:A Survey Image Source: https://cs.stanford.edu/~roozbeh/pascal-context/

The THUMOS14 (THUMOS 2014) dataset is a large-scale video dataset that includes 1,010 videos for validation and 1,574 videos for testing from 20 classes. Among all the videos, there are 220 and 212 videos with temporal annotations in validation and testing set, respectively. Source: Learning to Localize Actions from Moments Image Source: http://crcv.ucf.edu/THUMOS14/

The efforts to create a non-trivial and publicly available dataset for action recognition was initiated at the KTH Royal Institute of Technology in 2004. The KTH dataset is one of the most standard datasets, which contains six actions: walk, jog, run, box, hand-wave, and hand clap. To account for performance nuance, each action is performed by 25 different individuals, and the setting is systematically altered for each action per actor. Setting variations include: outdoor (s1), outdoor with scale variation (s2), outdoor with different clothes (s3), and indoor (s4). These variations test the ability of each algorithm to identify actions independent of the background, appearance of the actors, and the scale of the actors. Source: Review of Action Recognition and Detection Methods

Animals with Attributes (AwA) was a dataset for benchmarking transfer-learning algorithms, in particular attribute base classification. It consisted of 30475 images of 50 animals classes with six pre-extracted feature representations for each image. The animals classes are aligned with Osherson's classical class/attribute matrix, thereby providing 85 numeric attribute values for each class. Using the shared attributes, it is possible to transfer information between different classes. The Animals with Attributes dataset was suspended. Its images are not available anymore because of copyright restrictions. A drop-in replacement, Animals with Attributes 2, is available instead. Source: Transductive Multi-view Zero-Shot Learning Image Source: https://cvml.ist.ac.at/AwA/

The SNIPS Natural Language Understanding benchmark is a dataset of over 16,000 crowdsourced queries distributed among 7 user intents of various complexity: * SearchCreativeWork (e.g. Find me the I, Robot television show), * GetWeather (e.g. Is it windy in Boston, MA right now?), * BookRestaurant (e.g. I want to book a highly rated restaurant in Paris tomorrow night), * PlayMusic (e.g. Play the last track from Beyoncé off Spotify), * AddToPlaylist (e.g. Add Diamonds to my roadtrip playlist), * RateBook (e.g. Give 6 stars to Of Mice and Men), * SearchScreeningEvent (e.g. Check the showtimes for Wonder Woman in Paris). The training set contains of 13,084 utterances, the validation set and the test set contain 700 utterances each, with 100 queries per intent. Source: https://paperswithcode.com/paper/snips-voice-platform-an-embedded-spoken/

The ActivityNet Captions dataset is built on ActivityNet v1.3 which includes 20k YouTube untrimmed videos with 100k caption annotations. The videos are 120 seconds long on average. Most of the videos contain over 3 annotated events with corresponding start/end time and human-written sentences, which contain 13.5 words on average. The number of videos in train/validation/test split is 10024/4926/5044, respectively. Source: Bidirectional Attentive Fusion with Context Gating for Dense Video Captioning Image Source: https://cs.stanford.edu/people/ranjaykrishna/densevid/

OntoNotes 5.0 is a large corpus comprising various genres of text (news, conversational telephone speech, weblogs, usenet newsgroups, broadcast, talk shows) in three languages (English, Chinese, and Arabic) with structural information (syntax and predicate argument structure) and shallow semantics (word sense linked to an ontology and coreference). OntoNotes Release 5.0 contains the content of earlier releases - and adds source data from and/or additional annotations for, newswire, broadcast news, broadcast conversation, telephone conversation and web data in English and Chinese and newswire data in Arabic. Source: https://catalog.ldc.upenn.edu/LDC2013T19

JHMDB is an action recognition dataset that consists of 960 video sequences belonging to 21 actions. It is a subset of the larger HMDB51 dataset collected from digitized movies and YouTube videos. The dataset contains video and annotation for puppet flow per frame (approximated optimal flow on the person), puppet mask per frame, joint positions per frame, action label per clip and meta label per clip (camera motion, visible body parts, camera viewpoint, number of people, video quality). Source: Unsupervised Deep Metric Learning via Orthogonality based Probabilistic Loss Image Source: https://arxiv.org/pdf/1712.06316.pdf

Animals with Attributes 2 (AwA2) is a dataset for benchmarking transfer-learning algorithms, such as attribute base classification and zero-shot learning. AwA2 is a drop-in replacement of original Animals with Attributes (AwA) dataset, with more images released for each category. Specifically, AwA2 consists of in total 37322 images distributed in 50 animal categories. The AwA2 also provides a category-attribute matrix, which contains an 85-dim attribute vector (e.g., color, stripe, furry, size, and habitat) for each category. Source: Learning from Noisy Web Data with Category-level Supervision Image Source: https://arxiv.org/pdf/1604.00326.pdf

RESISC45 dataset is a dataset for Remote Sensing Image Scene Classification (RESISC). It contains 31,500 RGB images of size 256×256 divided into 45 scene classes, each class containing 700 images. Among its notable features, RESISC45 contains varying spatial resolution ranging from 20cm to more than 30m/px.

The UCY dataset consist of real pedestrian trajectories with rich multi-human interaction scenarios captured at 2.5 Hz (Δt=0.4s). It is composed of three sequences (Zara01, Zara02, and UCY), taken in public spaces from top-view. Source: Trajectron++: Dynamically-Feasible Trajectory Forecasting With Heterogeneous Data Image Source: http://trajnet.stanford.edu/data.php?n=1

This paper introduces the pipeline to scale the largest dataset in egocentric vision EPIC-KITCHENS. The effort culminates in EPIC-KITCHENS-100, a collection of 100 hours, 20M frames, 90K actions in 700 variable-length videos, capturing long-term unscripted activities in 45 environments, using head-mounted cameras. Compared to its previous version (EPIC-KITCHENS-55), EPIC-KITCHENS-100 has been annotated using a novel pipeline that allows denser (54% more actions per minute) and more complete annotations of fine-grained actions (+128% more action segments). This collection also enables evaluating the "test of time" - i.e. whether models trained on data collected in 2018 can generalise to new footage collected under the same hypotheses albeit "two years on". The dataset is aligned with 6 challenges: action recognition (full and weak supervision), action detection, action anticipation, cross-modal retrieval (from captions), as well as unsupervised domain adaptation for action recognition. For each challenge, we define the task, provide baselines and evaluation metrics.

aPY is a coarse-grained dataset composed of 15339 images from 3 broad categories (animals, objects and vehicles), further divided into a total of 32 subcategories (aeroplane, …, zebra). Source: From Classical to Generalized Zero-Shot Learning: a Simple Adaptation Process Image Source: https://www.cs.cmu.edu/~afarhadi/papers/Attributes.pdf