Papers with Code · CC BY-SA 4.0

데이터셋 탐색

데이터셋의 설명·홈페이지를 찾고, 관련 태스크의 평가 기록을 살펴보세요.

관련 논문 1편

Disaster

Disaster 설명 보기

Disaster is a dataset that contains images collected from various sources for three different disasters: fire, water and land. Besides this, it also contains images for various damaged infrastructure due to natural or man made calamities and damaged human due to war or accidents. There are 13,720 manually annotated images in this dataset, each image is annotated by three individuals. The authors are also providing discriminating image class information annotated manually with bounding box for a set of 200 test images. Images are collected from different news portals, social media, and standard datasets made available by other researchers.

관련 논문 8편

Human-Annotated Disaster Incidents Data

HumAID
HumAID 설명 보기

Social networks are widely used for information consumption and dissemination, especially during time-critical events such as natural disasters. Despite its significantly large volume, social media content is often too noisy for direct use in any application. Therefore, it is important to filter, categorize, and concisely summarize the available content to facilitate effective consumption and decision-making. To address such issues automatic classification systems have been developed using supervised modeling approaches, thanks to the earlier efforts on creating labeled datasets. However, existing datasets are limited in different aspects (e.g., size, contains duplicates) and less suitable to support more advanced and data-hungry deep learning models. HumAID is a large-scale dataset for crisis informatics research with ~77K human-labeled tweets, sampled from a pool of ~24 million tweets across 19 disaster events that happened between 2016 and 2019. The annotations in the provided datasets consists of following humanitarian categories. The dataset consists only english tweets and it is the largest dataset for crisis informatics so far. Humanitarian categories: * Caution and advice * Displaced people and evacuations * Don't know can't judge * Infrastructure and utility damage * Injured or dead people * Missing or found people * Not humanitarian * Other relevant information * Requests or urgent needs * Rescue volunteering or donation effort * Sympathy and support

관련 논문 3편

Tourism > Finance > Sales Revenue

Extreme Events > Natural Disasters > Hurricane
Extreme Events > Natural Disasters > Hurricane 설명 보기

A new spatio-temporal benchmark dataset (Hurricane), is suited for forecasting during extreme events and anomalies. The dataset is provided through the Florida Department of Revenue which provides the monthly sales revenue (2003-2020) for the tourism industry for all 67 counties of Florida which are prone to annual hurricanes. Furthermore, we aligned and joined the raw time series with the history of hurricane categories (i.e., event intensities) based on time for each county. Note that the hurricane category indicates the maximum sustained wind speed which can result in catastrophic damages as this number goes up (Category 1-6).

관련 논문 2편

Combination to Application

C2A: Human Detection in Disaster Scenarios
C2A: Human Detection in Disaster Scenarios 설명 보기

C2A: Combination to Application Dataset ## Overview This repository contains the code and information for the paper "UAV-Enhanced Combination to Application: Comprehensive Analysis and Benchmarking of a Human Detection Dataset for Disaster Scenarios" by Ragib Amin Nihal, Benjamin Yen, Katsutoshi Itoyama, and Kazuhiro Nakadai. The C2A (Combination to Application) dataset is a novel synthetic dataset designed to advance human detection in disaster scenarios using UAV imagery. It combines disaster scene backgrounds from the AIDER dataset with human poses from the LSP/MPII-MPHB dataset to create a comprehensive resource for training machine learning models. The full paper is available at: https://arxiv.org/pdf/2408.04922 ## Dataset The C2A dataset consists of 10,215 images containing over 360,000 annotated human instances in various disaster scenarios. It includes diverse human poses (bent, kneeling, lying, sitting, upright) and disaster contexts (traffic incidents, fire, flood, collapsed buildings). ### Usage Notes - The dataset is split into training, validation, and test sets. - Two annotation formats are provided for flexibility: YOLO and COCO. - Pose information is available in a separate folder for all images. - Users can choose between standard object detection (YOLO/COCO) or pose-aware detection (All labels with Pose info). ## Key Features - Synthetic dataset combining real disaster scenes with human poses - Over 360,000 annotated human instances - 5 human pose categories - 4 disaster scenario types - Image resolutions ranging from 123x152 to 5184x3456 pixels - Designed to improve human detection in complex disaster environments

관련 논문 2편

Titanic - Machine Learning from Disaster

Titanic
Titanic 설명 보기

Titanic Dataset Description ## Overview The data is divided into two groups: - Training set (train.csv): Used to build machine learning models. It includes the outcome (also called the "ground truth") for each passenger, allowing models to predict survival based on “features” like gender and class. Feature engineering can also be applied to create new features. - Test set (test.csv): Used to evaluate model performance on unseen data. The ground truth is not provided; the task is to predict survival for each passenger in the test set using the trained model. Additionally, gender_submission.csv is provided as an example submission file, containing predictions based on the assumption that all and only female passengers survive. --- ## Data Dictionary | Variable | Definition | Key | |------------|------------------------------------------|-------------------------------------------------| | survival | Survival | 0 = No, 1 = Yes | | pclass | Ticket class | 1 = 1st, 2 = 2nd, 3 = 3rd | | sex | Sex | | | age | Age in years | | | sibsp | # of siblings/spouses aboard the Titanic | | | parch | # of parents/children aboard the Titanic | | | ticket | Ticket number | | | fare | Passenger fare | | | cabin | Cabin number | | | embarked | Port of Embarkation | C = Cherbourg, Q = Queenstown, S = Southampton | --- ## Variable Notes - pclass: Proxy for socio-economic status (SES): - 1st = Upper - 2nd = Middle - 3rd = Lower - age: - Fractional if less than 1 year. - Estimated ages are represented in the form xx.5. - sibsp: Defines family relations as: - Sibling: Brother, sister, stepbrother, stepsister. - Spouse: Husband, wife (excluding mistresses and fiancés). - parch: Defines family relations as: - Parent: Mother, father. - Child: Daughter, son, stepdaughter, stepson. Some children traveled only with a nanny, so parch = 0 for them.

관련 논문 1편

Low Altitude Disaster Imagery v2

LADI v2
LADI v2 설명 보기

LADI Overview The Low Altitude Disaster Imagery (LADI) dataset was created to address the relative lack of annotated post-disaster aerial imagery in the computer vision community. Low altitude post-disaster aerial imagery from small planes and UAVs can provide high-resolution imagery to emergency management agencies to help them prioritize response efforts and perform damage assessments. In order to accelerate their workflow, computer vision can be used to automatically identify images that contain features of interest, including infrastructure such as buildings and roads, damage to such infrastructure, and hazards such as floods or debris. Version 1 of the LADI dataset was funded as part of a NIST public safety innovation accelerator effort to create datasets for public safety. It collected over 40k images from the Civil Air Patrol (CAP), which were annotated by crowdsourced workers providing labels from a hierarchical set of classes. The dataset was used in a number of NIST TRECVID challenges (2020, 2021, and 2022) and provided a broad basis for classification of aerial post-disaster scenes. However, the authors found that the quality of the labels were inconsistent, stemming in part from the untrained nature of the crowdsource workers, as well as the potentially subjective definition of certain labels, such as "damage". For LADI v2, the authors used CAP volunteers who were trained in the FEMA damage assessment process, and we collected damage labels using the defined FEMA Preliminary Damage Assessment scale: unaffected, affected, minor, major, destroyed. These damage levels have specific criteria, helping reduce the subjectivity of identifying whether a structure is damaged. We also provide pretrained classifiers to aid in replication, and serve as a basis for fine-tuning and potential deployments. ## Getting Started The LADI v2 dataset is available on Hugging Face at MITLL/LADI-v2-dataset. This is the recommended method. ### Default configuration The default configuration uses the v2a labels with images resized to fit within 1800x1200. This is the recommended configuration for most use cases. python from datasets import load_dataset ds = load_dataset("MITLL/LADI-v2-dataset") ### Advanced Usage If you need access to the full resolution images, the v2 label set, or the v1 dataset, you should load from the script revision. This will use a custom dataset loader script, which will require you to set revision=script and trust_remote_code=True. The available configurations for the script are: v2, v2a, v2_resized, v2a_resized, v1_damage, and v1_infra. The first time you load the dataset, you should pass download_ladi=True, which will download a local copy of the relevant dataset version to your local system at base_dir. python from datasets import load_dataset ds = load_dataset("MITLL/LADI-v2-dataset", "v2a_resized", revision="script", streaming=True, download_ladi=True, base_dir='./ladi_dataset', trust_remote_code=True) This only needs to be done once, and subsequent calls can omit the argument, and it will automatically use your local copy. python ds = load_dataset("MITLL/LADI-v2-dataset", "v2a_resized", revision="script", streaming=True, base_dir='./ladi_dataset', trust_remote_code=True) You can also manually access the dataset at https://ladi.s3.amazonaws.com/index.html, or via AWS S3 at s3://ladi, and the provided dataset class in training/LADI-v2-dataset. We recommend using the LADI_v2_resized version which resizes the images to 1800x1200, which should be large enough for most applications, but drastically reduces the overall dataset file size. The LADI v1 and v2 dataset files are hosted as part of the AWS Open Data program. ### Dataset Details The LADI-v2 dataset is a set of aerial disaster images captured and labeled by the Civil Air Patrol (CAP). The images are geotagged (in their EXIF metadata). Each image has been labeled in triplicate by CAP volunteers trained in the FEMA damage assessment process for multi-label classification; where volunteers disagreed about the presence of a class, a majority vote was taken. The classes are: - bridges_any - bridges_damage - buildings_affected - buildings_any - buildings_destroyed - buildings_major - buildings_minor - debris_any - flooding_any - flooding_structures - roads_any - roads_damage - trees_any - trees_damage - water_any The v2 dataset consists of approximately 10k images, split into a train set of 8k images, a validation set of 1k images, and a test test of 1k images. The train and validation sets are drawn from the same distribution (CAP images from federally-declared disasters 2015-2022), whereas the test set is drawn from events in 2023, which has a different distribution of event types and locations. This is done to simulate the distribution shift as new events occur each year. #### Dataset v2a The v2a variant of the dataset presents the same images with a modified set of labels, where the damage categories for buildings have been compressed into two classes of buildings_affected_or_greater and buildings_minor_or_greater. We find that this task is easier and of similar practical value for triage purposes. The bridges_damage label has also been removed due to the low number of positive examples in the dataset. - bridges_any - buildings_any - buildings_affected_or_greater - buildings_minor_or_greater - debris_any - flooding_any - flooding_structures - roads_any - roads_damage - trees_any - trees_damage - water_any #### Supported Tasks The images are labeled for multi-label classification, as any number of the elements listed above may be present in a single image. #### Data Structure A single example in the v2a dataset looks like this: { 'image': <PIL.PngImagePlugin.PngImageFile image mode=RGB size=1800x1200 at ...>, 'bridges_any': False, 'buildings_any': False, 'buildings_affected_or_greater': False, 'buildings_minor_or_greater': False, 'debris_any': False, 'flooding_any': False, 'flooding_structures': False, 'roads_any': False, 'roads_damage': False, 'trees_any': True, 'trees_damage': True, 'water_any': True } ### Pretrained Classifiers We provide a set of pretrained classifiers on the LADI v2 dataset for downstream finetuning and deployment purposes. See the associated model cards on Hugging Face for instructions on how to use. - LADI-v2-classifier-small - Recommended for deployment, based on google/bit-50, trained on entire LADI v2 dataset - LADI-v2-classifier-large - Based on microsoft/swinv2, trained on entire LADI v2 dataset We also provide "reference" versions of each model which are trained only on the train split of the LADI v2 dataset, to aid in reproducing the results presented in the paper. - LADI-v2-classifier-small-reference - LADI-v2-classifier-large-reference ## Citations ### LADI v2 If you use the LADI v2 dataset or classifiers, please cite the following: tex @misc{ladi_v2, title={LADI v2: Multi-label Dataset and Classifiers for Low-Altitude Disaster Imagery}, author={Samuel Scheele and Katherine Picchione and Jeffrey Liu}, year={2024}, eprint={2406.02780}, archivePrefix={arXiv}, primaryClass={cs.CV} } ## Distribution Information DISTRIBUTION STATEMENT A. Approved for public release. Distribution is unlimited. This material is based upon work supported by the Department of the Air Force under Air Force Contract No. FA8702-15-D-0001. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Department of the Air Force. © 2024 Massachusetts Institute of Technology. The software/firmware is provided to you on an As-Is basis Delivered to the U.S. Government with Unlimited Rights, as defined in DFARS Part 252.227-7013 or 7014 (Feb 2014). Notwithstanding any copyright notice, U.S. Government rights in this work are defined by DFARS 252.227-7013 or DFARS 252.227-7014 as detailed above. Use of this work other than as specifically authorized by the U.S. Government may violate any copyrights that exist in this work.