Papers with Code 아카이브

Data Augmentation

태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.

하위 태스크
2
평가표
3
보존된 평가 기록
26

Data augmentation involves techniques used for increasing the amount of data, based on different modifications, to expand the amount of examples in the original dataset. Data augmentation not only helps to grow the dataset but it also increases the diversity of the dataset. When training machine learning models, data augmentation acts as a regularizer and helps to avoid overfitting.

Data augmentation techniques have been found useful in domains like NLP and computer vision. In computer vision, transformations like cropping, flipping, and rotation are used. In NLP, data augmentation techniques can include swapping, deletion, random insertion, among others.

Further readings:

<span style="color:grey; opacity: 0.6">( Image credit: Albumentations )</span>

데이터셋

출처에 같은 태스크 이름으로 연결된 데이터셋입니다. 동명 태스크가 포함될 수 있으므로, 비교할 때는 평가 결과의 조건도 확인하세요.

Europarl-ASR

Europarl-ASR (EN) is a 1300-hour English-language speech and text corpus of parliamentary debates for (streaming) Automatic Speech Recognition training and benchmarking, speech data filtering and speech data verbatimization, based on European Parliament speeches and their official transcripts (1996-2020). Includes dev-test sets for streaming ASR benchmarking, made up of 18 hours of manually revised speeches. The availability of manual non-verbatim and verbatim transcripts for dev-test speeches makes this corpus also useful for the assessment of automatic filtering and verbatimization techniques. The corpus is released under an open licence at https://www.mllp.upv.es/europarl-asr/ Europarl-ASR CONTENTS: [Speech data] 1300 hours of English-language annotated speech data, 3 full sets of timed transcriptions (official non-verbatim, automatically noise-filtered, automatically verbatimized), 18 hours of speech data with both manually revised verbatim transcriptions and official non-verbatim transcriptions, split in 2 independent validation- evaluation partitions for 2 realistic ASR tasks (with vs. without previous knowledge of the speaker); [Text data] 70 million tokens of English-language text data; [Pretrained language models] the Europarl-ASR English-language n-gram language model and vocabulary.

Incidents

Contains 446,684 images annotated by humans that cover 43 incidents across a variety of scenes. Source: Detecting natural disasters, damage, and incidents in the wild

Hotels-50K

The Hotels-50K dataset consists of over 1 million images from 50,000 different hotels around the world. These images come from both travel websites, as well as the TraffickCam mobile application, which allows every day travelers to submit images of their hotel room in order to help combat trafficking. The TraffickCam images are more visually similar to images from trafficking investigations than the images from travel websites. The training dataset includes 1,027,871 images from 50,000 hotels, and 92 major hotel chains. Of the 50,000 hotels, 13,900 include user contributed images from the TraffickCam application (a total of 55,061 TraffickCam images are included in the training set). The test dataset includes 17,954 TraffickCam images from 5,000 different hotels (as well as versions of the test images that have medium and large occlusions to replicate the occlusions seen in real world trafficking victim photographs). Source: Hotels-50K: A Global Hotel Recognition Dataset

Intel Image Classification

Context This is image data of Natural Scenes around the world. Content This Data contains around 25k images of size 150x150 distributed under 6 categories. {'buildings' -> 0, 'forest' -> 1, 'glacier' -> 2, 'mountain' -> 3, 'sea' -> 4, 'street' -> 5 } The Train, Test and Prediction data is separated in each zip files. There are around 14k images in Train, 3k in Test and 7k in Prediction. This data was initially published on https://datahack.analyticsvidhya.com by Intel to host a Image classification Challenge. Acknowledgements Thanks to https://datahack.analyticsvidhya.com for the challenge and Intel for the Data Photo by Jan Böttinger on Unsplash Inspiration Want to build powerful Neural network that can classify these images with more accuracy.

m2caiSeg

Created from endoscopic video feeds of real-world surgical procedures. Overall, the data consists of 307 images, each of which is annotated for the organs and different surgical instruments present in the scene. Source: m2caiSeg: Semantic Segmentation of Laparoscopic Images using Convolutional Neural Networks

word2word

word2word contains easy-to-use word translations for 3,564 language pairs. - A large collection of freely & publicly available bilingual lexicons for 3,564 language pairs across 62 unique languages. - Easy-to-use Python interface for accessing top-k word translations and for building a new bilingual lexicon from a custom parallel corpus. - Constructed using a simple approach that yields bilingual lexicons with high coverage and competitive translation quality. Source: word2word

Amazon Fine Foods

Amazon Fine Foods is a dataset that consists of reviews of fine foods from amazon. The data span a period of more than 10 years, including all ~500,000 reviews up to October 2012. Reviews include product and user information, ratings, and a plaintext review. Source: https://snap.stanford.edu/data/web-FineFoods.html

Bentham

Bentham manuscripts refers to a large set of documents that were written by the renowned English philosopher and reformer Jeremy Bentham (1748-1832). Volunteers of the Transcribe Bentham initiative transcribed this collection. Currently, >6 000 documents or > 25 000 pages have been transcribed using this public web platform. For our experiments, we used the BenthamR0 dataset a part of the Bentham manuscripts.

BirdVox-full-night

The BirdVox-full-night dataset contains 6 audio recordings, each about ten hours in duration. These recordings come from ROBIN autonomous recording units, placed near Ithaca, NY, USA during the fall 2015. They were captured on the night of September 23rd, 2015, by six different sensors, originally numbered 1, 2, 3, 5, 7, and 10. Andrew Farnsworth used the Raven software to pinpoint every avian flight call in time and frequency. He found 35402 flight calls in total. He estimates that about 25 different species of passerines (thrushes, warblers, and sparrows) are present in this recording. Species are not labeled in BirdVox-full-night, but it is possible to tell apart thrushes from warblers and sparrrows by looking at the center frequencies of their calls. The annotation process took 102 hours. Source: https://wp.nyu.edu/birdvox/birdvox-full-night/ Image Source: https://wp.nyu.edu/birdvox/birdvox-full-night/

Konzil

Konzil dataset was created by specialists of the University of Greifswald. It contains manuscripts written in modern German. Train sample consists of 353 lines, validation - 29 lines and test - 87 lines.

LITIS Rouen

The LITIS-Rouen dataset is a dataset for audio scenes. It consists of 3026 examples of 19 scene categories. Each class is specific to a location such as a train station or an open market. The audio recordings have a duration of 30 seconds and a sampling rate of 22050 Hz. The dataset has a total duration of 1500 minutes. Source: Spatio-Temporal Attention Pooling for Audio Scene Classification Image Source: https://www.researchgate.net/figure/Summary-of-Litis-Rouen-audio-scene-dataset_tbl1_329608235

Patzig

Patzig contains handwritten texts written in modern German. Train sample consists of 485 lines, validation - 38 lines and test -118 lines.

Ricordi

Ricordi contains handwritten texts written in Italian. Train sample consists of 295 lines, validation - 19 lines and test - 69 lines.

Schiller

Schiller contains handwritten texts written in modern German. Train sample consists of 244 lines, validation - 21 lines and test - 63 lines.

Schwerin

Schwerin contains handwritten texts written in medieval German. Train sample consists of 793 lines, validation - 68 lines and test - 196 lines.

ELFW

Extended Labeled Faces in-the-Wild (ELFW) is a dataset supplementing with additional face-related categories —and also additional faces— the originally released semantic labels in the vastly used Labeled Faces in-the-Wild (LFW) dataset. Additionally, two object-based data augmentation techniques are deployed to synthetically enrich under-represented categories which, in benchmarking experiments, reveal that not only segmenting the augmented categories improves, but also the remaining ones benefit. Source: Extended Labeled Faces in-the-Wild Image Source: https://multimedia-eurecat.github.io/2020/06/22/extended-faces-in-the-wild.html

MVTec D2S

MVTec D2S is a benchmark for instance-aware semantic segmentation in an industrial domain. It contains 21,000 high-resolution images with pixel-wise labels of all object instances. The objects comprise groceries and everyday products from 60 categories. The benchmark is designed such that it resembles the real-world setting of an automatic checkout, inventory, or warehouse system. The training images only contain objects of a single class on a homogeneous background, while the validation and test sets are much more complex and diverse. Source: MVTec D2S: Densely Segmented Supermarket Dataset Image Source: https://www.mvtec.com/company/research/datasets/mvtec-d2s

Saint Gall

Saint Gall dataset contains handwritten historical manuscripts written in Latin that date back to the 9th century. It consists of 60 pages, 1 410 text lines and 11 597 words.

UFPR-AMR

This dataset contains 2,000 images taken from inside a warehouse of the Energy Company of Paraná (Copel), which directly serves more than 4 million consuming units in the Brazilian state of Paraná. The images were acquired with three different cameras and are available in the JPG format with a resolution between 2,340 × 4,160 and 3,120 × 4,160 pixels. The dataset is split into three sets: training (800 images), validation (400 images) and testing (800 images). Every image has the following annotations available in a text file: the camera in which the image was taken, the counter’s position (x,y,w,h) and reading, as well as the position of each digit. All counters of the dataset (regardless of meter type) have 5 digits, and thus 10,000 digits were manually annotated. Source: Convolutional Neural Networks for Automatic Meter Reading

CECW

The CECW dataset is a color-extended version of the Cleanup World (CW) borrowed from the mobile-manipulation robot domain. CW refers to a world equipped with a movable object as well as four rooms in four colors, including "blue," "green," "red," and "yellow," which is designed as a simulation environment where the agent can act based on the instructions received. CW obeys a particular Geometric Linear Temporal Logic (GLTL) to parse commands by grammatical syntax, resulting in a total of 3,382 commands reflecting 39 GLTL expressions. Source: https://github.com/MrShininnnnn/CECW Image Source: https://github.com/MrShininnnnn/CECW

COCO Earthquake

COCO Earthquake is a dataset similar to Common Objects in Context (COCO) used for cracking segmentation. The images selected in the dataset are at various scales, and the tool referred to as the COCO Annotator is used to label cracks for training. In these labeled images, cracks are in yellow and background is in purple. Size of the training and labeling images is varied from 168×300 to 4600×3070. By excluding steel structures, 2,021 images are labeled when surface cracks appeared on structural or nonstructural materials at various scales.

ImagiFilter

ImagiFilter focusses on photographic and/or natural images, a very common use-case in computer vision research. Annotations for coarse prediction are provided, i.e. photographic vs. non-photographic, and smaller fine-grained prediction tasks where the non-photographic class is broken down into five classes: maps, drawings, graphs, icons, and sketches. Source: ImagiFilter: A resource to enable the semi-automatic mining of images at scale

RadioTalk

RadioTalk is a corpus of speech recognition transcripts sampled from talk radio broadcasts in the United States between October of 2018 and March of 2019. The corpus is intended for use by researchers in the fields of natural language processing, conversational analysis, and the social sciences. The corpus encompasses approximately 2.8 billion words of automatically transcribed speech from 284,000 hours of radio, together with metadata about the speech, such as geographical location, speaker turn boundaries, gender, and radio program information. Source: https://github.com/social-machines/RadioTalk

Synthetic COVID-19 CXR Dataset

A public open dataset of synthetic chest X-ray images of COVID-19. The dataset consists of 21,295 synthetic COVID-19 chest X-ray images. Images are generated using an unsupervised domain adaptation approach by leveraging class conditioning and adversarial training from source datasets RSNA Kaggle Dataset and COVID-19 Image Data Collection. Implementation of the algorithm is available here.

Tilt-RGBD

Includes considerable roll and pitch camera motion. Source: Surface Normal Estimation of Tilted Images via Spatial Rectifier

출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.