From 3ebd70a873ed0ccc96e186c11a35f59b92afeda5 Mon Sep 17 00:00:00 2001 From: caizhongang Date: Fri, 29 Mar 2024 22:49:56 +0800 Subject: [PATCH] Release the updated version of SMPLer-X-H32 --- README.md | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/README.md b/README.md index 2d1625e..4ee1aac 100644 --- a/README.md +++ b/README.md @@ -13,6 +13,7 @@ ## News +- [2024-03-29] An updated version of SMPLer-X-H32 is released to fix camera estimation on 3DPW-like data. - [2024-02-29] [HuggingFace](https://huggingface.co/spaces/caizhongang/SMPLer-X) demo is online! - [2023-10-23] Support visualization through SMPL-X mesh overlay and add inference docker. - [2023-10-02] [arXiv](https://arxiv.org/abs/2309.17448) preprint is online! @@ -56,14 +57,16 @@ docker run --gpus all -v :/smplerx_inference/vid_input \ ## Pretrained Models -| Model | Backbone | #Datasets | #Inst. | #Params | MPE | Download | FPS | -|:------------:|:--------:|:---------:|:------:|:-------:|:----:|:--------:|:-----:| -| SMPLer-X-S32 | ViT-S | 32 | 4.5M | 32M | 82.6 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EbkyKOS5PclHtDSxdZDmsu0BNviaTKUbF5QUPJ08hfKuKg?e=LQVvzs) | 36.17 | -| SMPLer-X-B32 | ViT-B | 32 | 4.5M | 103M | 74.3 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EVcRBwNOQl9OtWhnCU54l58BzJaYEPxcFIw7u_GnnlPZiA?e=nPqMjz) | 33.09 | -| SMPLer-X-L32 | ViT-L | 32 | 4.5M | 327M | 66.2 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EWypJXfmJ2dEhoC0pHFFd5MBoSs7LCZmWQjHjbcQJF72fQ?e=Gteus3) | 24.44 | -| SMPLer-X-H32 | ViT-H | 32 | 4.5M | 662M | 63.0 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/Eco7AAc_ZmtBrhAat2e5Ti8BonrR3NVNx-tNSck45ixT4Q?e=nudXrR) | 17.47 | +| Model | Backbone | #Datasets | #Inst. | #Params | MPE | Download | FPS | +|:-------------:|:--------:|:---------:|:------:|:-------:|:----:|:--------:|:-----:| +| SMPLer-X-S32 | ViT-S | 32 | 4.5M | 32M | 82.6 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EbkyKOS5PclHtDSxdZDmsu0BNviaTKUbF5QUPJ08hfKuKg?e=LQVvzs) | 36.17 | +| SMPLer-X-B32 | ViT-B | 32 | 4.5M | 103M | 74.3 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EVcRBwNOQl9OtWhnCU54l58BzJaYEPxcFIw7u_GnnlPZiA?e=nPqMjz) | 33.09 | +| SMPLer-X-L32 | ViT-L | 32 | 4.5M | 327M | 66.2 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EWypJXfmJ2dEhoC0pHFFd5MBoSs7LCZmWQjHjbcQJF72fQ?e=Gteus3) | 24.44 | +| SMPLer-X-H32 | ViT-H | 32 | 4.5M | 662M | 63.0 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/Eco7AAc_ZmtBrhAat2e5Ti8BonrR3NVNx-tNSck45ixT4Q?e=nudXrR) | 17.47 | +| SMPLer-X-H32* | ViT-H | 32 | 4.5M | 662M | 59.7 | [model](https://pjlab-my.sharepoint.cn/:u:/g/personal/openmmlab_pjlab_org_cn/EYdhbS9ZnctPgIs5PZNZBaIBVMJUr4LKBWCIUuBBVmbwAg?e=kLDj5P) | 17.47 | * MPE (Mean Primary Error): the average of the primary errors on five benchmarks (AGORA, EgoBody, UBody, 3DPW, and EHF) * FPS (Frames Per Second): the average inference speed on a single Tesla V100 GPU, batch size = 1 +* SMPLer-X-H32* is the updated version of SMPLer-X-H32, which fixes the camera estimation issue on 3DPW-like data. ## Preparation - download all datasets