Papers with Code · CC BY-SA 4.0

데이터셋 탐색

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

관련 논문 31편

SUN Database

SUN
SUN 설명 보기

When glancing at a magazine, or browsing the Internet, we are continuously being exposed to photographs. Despite of this overflow of visual information, humans are extremely good at remembering thousands of pictures along with some of their visual details. But not all images are equal in memory. Some stitch to our minds, and other are forgotten. In this paper we focus on the problem of predicting how memorable an image will be. We show that memorability is a stable property of an image that is shared across different viewers. We introduce a database for which we have measured the probability that each picture will be remembered after a single view. We analyze image features and labels that contribute to making an image memorable, and we train a predictor based on global image descriptors. We find that predicting image memorability is a task that can be addressed with current computer vision techniques. Whereas making memorable images is a challenging task in visualization and photography, this work is a first attempt to quantify this useful quality of images.

관련 논문 867편

Large-scale Scene UNderstanding Challenge

LSUN
LSUN 설명 보기

The Large-scale Scene Understanding (LSUN) challenge aims to provide a different benchmark for large-scale scene classification and understanding. The LSUN classification dataset contains 10 scene categories, such as dining room, bedroom, chicken, outdoor church, and so on. For training data, each category contains a huge number of images, ranging from around 120,000 to 3,000,000. The validation data includes 300 images, and the test data has 1000 images for each category. Source: Knowledge Guided Disambiguation for Large-Scale Scene Classification with Multi-Resolution CNNs Image Source: https://www.yf.io/p/lsun

관련 논문 477편

SUN RGB-D

SUN RGB-D 설명 보기

The SUN RGBD dataset contains 10335 real RGB-D images of room scenes. Each RGB image has a corresponding depth and segmentation map. As many as 700 object categories are labeled. The training and testing sets contain 5285 and 5050 images, respectively. Source: Mix and match networks: multi-domain alignment for unpaired image-to-image translation Image Source: https://rgbd.cs.princeton.edu/

관련 논문 126편

SUN3D

SUN3D 설명 보기

SUN3D contains a large-scale RGB-D video database, with 8 annotated sequences. Each frame has a semantic segmentation of the objects in the scene and information about the camera pose. It is composed by 415 sequences captured in 254 different spaces, in 41 different buildings. Moreover, some places have been captured multiple times at different moments of the day. Source: A Review on Deep Learning TechniquesApplied to Semantic Segmentation Image Source: http://sun3d.cs.princeton.edu/

관련 논문 93편

Scene UNderstanding 360° panorama

SUN360
SUN360 설명 보기

The goal of the SUN360 panorama database is to provide academic researchers in computer vision, computer graphics and computational photography, cognition and neuroscience, human perception, machine learning and data mining, with a comprehensive collection of annotated panoramas covering 360x180-degree full view for a large variety of environmental scenes, places and the objects within. To build the core of the dataset, the authors download a huge number of high-resolution panorama images from the Internet, and group them into different place categories. Then, they designed a WebGL annotation tool for annotating the polygons and cuboids for objects in the scene. Source: Scene UNderstanding 360° panorama Image Source: http://3dvision.princeton.edu/projects/2012/SUN360/

관련 논문 38편

SUN Attribute

SUN Attribute 설명 보기

The SUN Attribute dataset consists of 14,340 images from 717 scene categories, and each category is annotated with a taxonomy of 102 discriminate attributes. The dataset can be used for high-level scene understanding and fine-grained scene recognition. Source: Zero-Shot Learning with Multi-Battery Factor Analysis Image Source: https://cs.brown.edu/~gmpatter/sunattributes.html

관련 논문 26편

SUN09

SUN09 설명 보기

The SUN09 dataset consists of 12,000 annotated images with more than 200 object categories. It consists of natural, indoor and outdoor images. Each image contains an average of 7 different annotated objects and the average occupancy of each object is 5% of image size. The frequencies of object categories follow a power law distribution. Source: A Pooling Approach to Modelling Spatial Relations forImage Retrieval and Annotation Image Source: http://people.csail.mit.edu/myungjin/HContext.html

관련 논문 19편

ImageNet-1k vs SUN

ImageNet-1k vs SUN 설명 보기

A benchmark dataset for out-of-distribution detection. ImageNet-1k is in-distribution, while SUN is out-of-distribution.

관련 논문 17편

SUN-SEG-Easy (Unseen)

SUN-SEG-Easy (Unseen) 설명 보기

The SUN-SEG dataset is a high-quality per-frame annotated VPS dataset, which includes 158,690 frames from the famous SUN dataset. It extends the labels with diverse types, i.e., object mask, boundary, scribble, polygon, and visual attribute. It also introduces the pathological information from the original SUN dataset, including pathological classification labels, location information, and shape information. Notably, the origin SUN dataset has 113 colonoscopy videos, including 100 positive cases with 49, 136 polyp frames and 13 negative cases with 109, 554 non-polyp frames. It manually trims them into 378 positive and 728 negative short clips, meanwhile maintaining their intrinsic consecutive relationship. Such data pre-processing ensures each clip has around 3~11s duration at a real-time frame rate (i.e., 30 fps), which promotes the fault-tolerant margin for various algorithms and devices. To this end, the re-organized SUN-SEG contains 1, 106 short video clips with 158, 690 video frames totally, offering a solid foundation to build a representative benchmark. As such, it yields the final version of our SUN-SEG dataset, which includes 49,136 polyp frames (i.e., positive part) and 109,554 non-polyp frames (i.e., negative part) taken from different 285 and 728 colonoscopy videos clips, as well as the corresponding annotations.

관련 논문 17편

SUN-SEG-Hard (Unseen)

SUN-SEG-Hard (Unseen) 설명 보기

The SUN-SEG dataset is a high-quality per-frame annotated VPS dataset, which includes 158,690 frames from the famous SUN dataset. It extends the labels with diverse types, i.e., object mask, boundary, scribble, polygon, and visual attribute. It also introduces the pathological information from the original SUN dataset, including pathological classification labels, location information, and shape information. Notably, the origin SUN dataset has 113 colonoscopy videos, including 100 positive cases with 49, 136 polyp frames and 13 negative cases with 109, 554 non-polyp frames. It manually trims them into 378 positive and 728 negative short clips, meanwhile maintaining their intrinsic consecutive relationship. Such data pre-processing ensures each clip has around 3~11s duration at a real-time frame rate (i.e., 30 fps), which promotes the fault-tolerant margin for various algorithms and devices. To this end, the re-organized SUN-SEG contains 1, 106 short video clips with 158, 690 video frames totally, offering a solid foundation to build a representative benchmark. As such, it yields the final version of our SUN-SEG dataset, which includes 49,136 polyp frames (i.e., positive part) and 109,554 non-polyp frames (i.e., negative part) taken from different 285 and 728 colonoscopy videos clips, as well as the corresponding annotations.

관련 논문 11편

The Samsung Open MOS Dataset for the Evaluation of Neural Text-to-Speech Synthesis

SOMOS
SOMOS 설명 보기

The SOMOS dataset is a large-scale mean opinion scores (MOS) dataset consisting of solely neural text-to-speech (TTS) samples. It can be employed to train automatic MOS prediction systems focused on the assessment of modern synthesizers, and can stimulate advancements in acoustic model evaluation. It consists of 20K synthetic utterances of the LJ Speech voice, a public domain speech dataset which is a common benchmark for building neural acoustic models and vocoders. Utterances are generated from 200 TTS systems including vanilla neural acoustic models as well as models which allow prosodic variations.

관련 논문 7편

Sunnybrook Cardiac Data

Sunnybrook Cardiac Data 설명 보기

The Sunnybrook Cardiac Data (SCD), also known as the 2009 Cardiac MR Left Ventricle Segmentation Challenge data, consist of 45 cine-MRI images from a mixed of patients and pathologies: healthy, hypertrophy, heart failure with infarction and heart failure without infarction. Subset of this data set was first used in the automated myocardium segmentation challenge from short-axis MRI, held by a MICCAI workshop in 2009. The whole complete data set is now available in the CAP database with public domain license. There are four pathological groups in this data set, which were classified based on (K Alfakih et al., JMRI 2003) paper, i.e.: Heart failure with infarction (HF-I) group had ejection fraction (EF) < 40% and evidence of late gadolinium (Gd) enhancement. Heart failure without infarction (HF) group had EF < 40% and no late Gd enhancement. LV hypertrophy (HYP) group had normal EF (> 55%) and a ratio of left ventricular (LV) mass over body surface area is > 83 g/m2. Healthy (N) group had EF > 55% and no hypertrophy.

관련 논문 4편

Kitsune Network Attack Dataset

Kitsune Network Attack Dataset 설명 보기

Kitsune Network Attack Dataset This is a collection of nine network attack datasets captured from a either an IP-based commercial surveillance system or a network full of IoT devices. Each dataset contains millions of network packets and diffrent cyber attack within it. For each attack, you are supplied with: * A preprocessed dataset in csv format (ready for machine learning) * The corresponding label vector in csv format * The original network capture in pcap format (in case you want to engineer your own features) We will now describe in detail what's in these datasets and how they were collected. ### The Network Attacks We have collected a wide variety of attacks which you would find in a real network intrusion. The following is a list of the cyber attack datasets avalaible: <img src="https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F827271%2F79e305668553e521b0709a2413323c45%2Fkaggle_dataset_table.png?generation=1598461684070844&alt=media" alt="image" width="100"> For more details on the attacks themselves, please refer to our NDSS paper (citation below). The Data Collection The following figure presents the network topologies which we used to collect the data, and the corrisponding attack vectors at which the attacks were performed. The network capture took place at point 1 and point X at the router (where a network intrusion detection system could feasibly be placed). For each dataset, clean network traffic was captured for the first 1 million packets, then the cyber attack was performed. ### The Dataset Format Each preprocessed dataset csv has m rows (packets) and 115 columns (features) with no header. The 115 features were extracted using our AfterImage feature extractor, described in our NDSS paper (see below) and available in Python here. In summary, the 115 features provide a statistical snapshot of the network (hosts and behaviors) in the context of the current packet traversing the network. The AfterImage feature extractor is unique in that it can efficiently process millions of streams (network channels) in real-time, incrementally, making it suitable for handling network traffic. Citation If you use these datasets, please cite: @inproceedings{mirsky2018kitsune, title={Kitsune: An Ensemble of Autoencoders for Online Network Intrusion Detection}, author={Mirsky, Yisroel and Doitshman, Tomer and Elovici, Yuval and Shabtai, Asaf}, booktitle={The Network and Distributed System Security Symposium (NDSS) 2018}, year={2018} }

관련 논문 2편

TBSI Sunwoda Battery Dataset

TBSI Sunwoda Battery Dataset 설명 보기

AI for science has generated a great deal of enthusiasm from both academia and industry. The field of battery energy storage is no exception due to its cross-cutting properties of materials, chemistry, physics and electrical engineering. Due to the complexity and uncertainty of the manufacturing process, there persistently exists a considerable mismatch in performance between a manufactured battery and its counterpart from material laboratory, leading to compromised product quality, R&D efficiency, investment cost and lifetime sustainability. Sunwoda Electronic Co., Ltd, generates the TBSI Sunwoda Battery Dataset to verify the performance of novel battery material composition designs. The collaboration team at Tsinghua Berkeley Shenzhen Institute (TBSI) performs the main research work by providing an efficient and reliable early battery prototype verification methodology. We open-source this dataset to inspire more diversified data-driven, physics-informed battery management research and real-world applications, including, but not limited to, state of charge (SOC) estimation, state of health (SOH) estimation, remaining useful life (RUL) prediction, degradation trajectory prediction, consistency management, and thermal management.

관련 논문 1편

RT-Percept Sun Temple

RT-Percept Sun Temple 설명 보기

Pre-rendered dataset used in Training and Predicting Visual Error for Real-Time Applications for the Sun Temple scene. Generated using the RT-Percept renderer and the RT-Percept scenes.

관련 논문 1편

SUN-RGBD-IS

SUN-RGBD-IS 설명 보기

A RGB-D dataset converted from SUN-RGBD into COCO-style instance segmentation format. To transform SUN-RGBD into an instance segmentation benchmark (i.e., SUN-RGBDIS), we employed a pipeline similar to that of NYUDv2-IS. We selected 17 categories from the original 37 classes, carefully omitting non-instance categories like ceilings and walls. Images lacking any identifiable object instances were filtered out to maintain dataset relevance for instance segmentation tasks. We systematically convert segmentation annotations into COCO format, generating precise bounding boxes, instance masks, and object attributes.

관련 논문 1편

SVOX Sun

SVOX Sun 설명 보기

Click to add a brief description of the dataset (Markdown and LaTeX enabled). Provide: * a high-level explanation of the dataset characteristics * explain motivations and summary of its content * potential use cases of the dataset

관련 논문 1편

Sentinel 2 manually extracted deep water spectra with high noise levels and sunglint

Sentinel 2 manually extracted deep water spectra with high noise levels and sunglint 설명 보기

This dataset includes 2.133.324 reflectance water spectra which were manually extracted by visual observation from 30 Sentinel 2 level 1C satellite images. The spectra were extracted from deep water areas with high noise levels and sunglint. The Sentinel 2 images depicted 2 tiles of the same orbit and were collected in 2016 (2 images), 2017 (19 images) and 2018 (9 images). The images contain 13 bands, 3 with 60 m spatial resolution, 4 with 10 m spatial resolution and 6 with 20 m spatial resolution. Before the spectra extraction, the bands with spatial resolution 10 and 20 m were resampled to 60 m and then the images were cropped in order to remove the land and depict optically homogenous sea regions. A figure depicting the location of the Sentinel 2 tiles (white polygons (1,2)) and the cropped tiles (red polygons (3,4)) is included in this folder. A figure depicting example scenes from which spectra were obtained through regions of interest (rois) is included as well. The spectra are stored in .csv files. Each file is named after the name of the Sentinel 2 product which includes sensing and creation date as well the relative orbit number and tile code. The content of each file includes latitude and longitude coordinates (UTM/WGS84 projection) of each spectral signature as well as the reflectance values of the 13 Sentinel 2 bands. This dataset was created for the purpose of the study described in https://www.tandfonline.com/doi/full/10.1080/01431161.2020.1714776

관련 논문 1편

SunspotsYoloDataset: annotated solar images captured with smart telescopes (January 2023 - May 2024)

SunspotsYoloDataset: annotated solar images captured with smart telescopes (January 2023 - May 2024) 설명 보기

SunspotsYoloDataset is a set of 1690+380+128 high-resolution RGB astronomical images captured with smart telescopes with specific solar filters and annotated with the positions of sunspots that are effectively in the images. Two instruments were used for several months from Luxembourg and France between January 2023 and May 2024: a Stellina smart telescope (https://vaonis.com/stellina) and a Vespera smart telescope (https://vaonis.com/vespera). SunspotsYoloDataset can be used to train YOLO detection models on solar images, enabling the prediction of unexpected events such as Borealis Aurora with astronomical equipment accessible to the public. SunspotsYoloDataset is formatted with the YOLO standard, i.e., with separated files for images and annotations, usable by state-of-the-art training tools and graphical software like MakeSense (https://www.makesense.ai). More precisely, there is a ZIP file containing RGB images in JPEG format (minimal compression), and text files containing the positions of sunspots. Each RGB image has a resolution of 640 × 640 pixels. For more details about the dataset, please contact the author: olivier.parisot@list.lu . For more information about Luxembourg of Science and Technology (LIST), please consult: https://www.list.lu .