No. | Figure | Title | features | Results | Pub. | Links |
---|---|---|---|---|---|---|
1 | All for One: Frame-wise Rank Loss for Improving Video-based Person Re-identification | Resnet50 + GRU + Rankloss | PRID(Rank1=75.17%) Mars(Rank1=77.27%,mAP=64.76%) | ICASSP2019 | paper | |
2 | Spatial-Temporal Attention-aware Learning for Video-based Person Re-identification | GoogleNet + STAL | iLIDS-VID(Rank1=82.8%) PRID(Rank1=92.7%) Mars(Rank1=82.2%,mAP=73.5%) | ITIP2019 | paper | |
3 | Joint Attentive Spatial-Temporal Feature Aggregation for Video-Based Person Re-Identification | FCN + attention | iLIDS-VID(Rank1=76%) PRID(Rank1=97.4%) Mars(Rank1=89.3%,mAP=74.9%) | IEEE Access | paper | |
4 | k-Reciprocal Harmonious Attention Network for Video-Based Person Re-Identification | Resnet50 + Attention | iLIDS-VID(Rank1=80.3%) PRID(Rank1=90.0%) Mars(Rank1=84.9%,mAp=76.7%) | IEEE Access | paper | |
5 | Intra-clip Aggregation for Video Person Re-identification | Resnet50 + Synchronized Transformation (ST) and Intra-clip Aggregation (ICA) | iLIDS-VID(Rank1=88.7%) Mars(Rank1=86.0%,mAP=80.8%) | Arxiv 2019 | paper | |
6 | Video-based Person Re-identification with Two-stream Convolutional Network and Co-attentive Snippet Embedding | Resnet50 + Attention + Pose and Optical map | iLIDS-VID(Rank1=88.7%) PRID(Rank1=94.4%) | Arxiv 2019 | paper | |
7 | Attribute-Driven Feature Disentangling and Temporal Aggregation for Video Person Re-Identification | Resnet50 + attribute learning | iLIDS-VID(Rank1=86.3%) PRID(Rank1=93.9%) Mars(Rank1=82.6%,mAP=71.2%) | CVPR2019 | paper | |
8 | VRSTC: Occlusion-Free Video Person Re-Identification | Resnet50 + Attention + GAN | iLIDS-VID(Rank1=83.4%) Mars(Rank1=88.5%,mAP=82.3%) DukeMTMC(Rank1=95.0%,mAP=93.5%) | CVPR2019 | paper | |
9 | Spatially and Temporally Efficient Non-local Attention Network for Video-based Person Re-Identification | Resnet50 + Non-Local | Mars(Rank1=90%,mAP=82.8%) DukeMTMC(Rank1=96.3%,mAP=94.9%) | BMVC2019 | paper code | |
10 | Global-Local Temporal Representations For Video Person Re-Identification | Resnet50 + parallel dilated convolutions + temporal self-attention | Mars(Rank1=87.02%,mAP=78.47%) DukeMTMC(Rank1=96.29%,mAP=93.74%) iLIDS-VID(Rank1=86.0%) PRID(Rank1=95.5%) LS-VID(Rank1=63.07%,mAP=44.32%) | ICCV2019 | paper | |
11 | Adaptive Graph Representation Learning for Video Person Re-identification | Resnet50 + pose alignment + GNN | Mars(Rank1=89.8%,mAP=81.1%) DukeMTMC(Rank1=97.0%,mAP=95.4%) iLIDS-VID(Rank1=84.5%) PRID(Rank1=94.6%) | IEEE TIP | paper | |
12 | Co-Segmentation Inspired Attention Networks for Video-Based Person Re-Identification | SE-Resnet50 + Cosegmentation based Attention Module(COSAM) | Mars(Rank1=84.9%,mAP=79.9%) DukeMTMC(Rank1=95.4%,mAP=94.1%) iLIDS-VID(Rank1=79.61%) | ICCV2019 | paper | |
13 | STA: Spatial-Temporal Attention for Large-Scale Video-based Person Re-Identification | Resnet50 + Local Attention | Mars(Rank1=86.3%,mAP=80.8%) DukeMTMC(Rank1=96.2%,mAP=94.9%) | AAAI2019 | paper |
Files
2019
Folders and files
Name | Name | Last commit date | ||
---|---|---|---|---|
parent directory.. | ||||