Papers with Code · CC BY-SA 4.0

데이터셋 탐색

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

관련 논문 69편

Dongguk Body-based Person Recognition Database (DBPerson-Recog-DB1)

RegDB
RegDB 설명 보기

RegDB is used for Visible-Infrared Re-ID which handles the cross-modality matching between the daytime visible and night-time infrared images. The dataset contains images of 412 people. It includes 10 color and 10 thermal images for each person. Source: Person Recognition System Based on a Combination of Body Images from Visible Light and Thermal Cameras

관련 논문 4편

RegDB-C

RegDB-C 설명 보기

RegDB-C is an evaluation set that consists of algorithmically generated corruptions applied to the RegDB test-set (color images). These corruptions consist of Noise: Gaussian, shot, impulse, and speckle; Blur: defocus, frosted glass, motion, zoom, and Gaussian; Weather: snow, frost, fog, brightness, spatter, and rain; Digital: contrast, elastic, pixel, JPEG compression, and saturate. Each corruption has five severity levels, resulting in 100 distinct corruptions.

관련 논문 1편

RegDB-C*

RegDB-C* 설명 보기

RegDB-C* is an evaluation set that consists of algorithmically generated corruptions applied to the RegDB test-set, and especially to both the visible and the thermal data. In comparison with the RegDB-C dataset proposed by Chen et al. in "Benchmarks for Corruption Invariant Person Re-identification" paper, our dataset is used in a multimodal manner and do not consider visible data corruptions only. Used corruptions are globally the same; Noise: Gaussian, shot, impulse, and speckle; Blur: defocus, frosted glass, motion, zoom, and Gaussian; Weather: snow, frost, fog, brightness, spatter, and rain; Digital: contrast, elastic, pixel, JPEG compression, and saturate. However, corruptions are adapted to respect the thermal modality encoding, and brightness is not used to corrupt the thermal data. Five severity levels are considered per corruption.