CPF: Learning a Contact Potential Field to Model the Hand-object Interaction

Last update: Dec 26, 2022

Related tags

Overview

Contact Potential Field

This repo contains model, demo, and test codes of our paper: CPF: Learning a Contact Potential Field to Model the Hand-object Interaction

Guide to the Demo

1. Get our code:

$ git clone --recursive https://github.com/lixiny/CPF.git
$ cd CPF

2. Set up your new environment:

$ conda env create -f environment.yaml
$ conda activate cpf

3. Download assets files and put it in `assets` folder.

Download the MANO model files from official MANO website, and put it into assets/mano. We currently only use the MANO_RIGHT.pkl

Now your assets folder should look like this:

.
├── anchor/
│   ├── anchor_mapping_path.pkl
│   ├── anchor_weight.txt
│   ├── face_vertex_idx.txt
│   └── merged_vertex_assignment.txt
├── closed_hand/
│   └── hand_mesh_close.obj
├── fhbhands_fits/
│   ├── Subject_1/
│   │   ├── ...
│   ├── Subject_2/
|   ├── ...
├── hand_palm_full.txt
└── mano/
    ├── fhb_skel_centeridx9.pkl
    ├── info.txt
    ├── LICENSE.txt
    └── MANO_RIGHT.pkl

4. Download Dataset

First-Person Hand Action Benchmark (fhb)

Download and unzip the First-Person Hand Action Benchmark dataset following the official instructions to the data/fhbhands folder If everything is correct, your data/fhbhands should look like this:

.
├── action_object_info.txt
├── action_sequences_normalized/
├── change_log.txt
├── data_split_action_recognition.txt
├── file_system.jpg
├── Hand_pose_annotation_v1/
├── Object_6D_pose_annotation_v1_1/
├── Object_models/
├── Subjects_info/
├── Video_files/
├── Video_files_480/ # Optionally

Optionally, resize the images (speeds up training !) based on the handobjectconsist/reduce_fphab.py.

$ python reduce_fphab.py

Download our fhbhands_supp and place it at data/fhbhands_supp:

Download our fhbhands_example and place it at data/fhbhands_example. This fhbhands_example contains 10 samples that are designed to demonstrate our pipeline.

├── fhbhands/
├── fhbhands_supp/
│   ├── Object_models/
│   └── Object_models_binvox/
├── fhbhands_example/
│   ├── annotations/
│   ├── images/
│   ├── object_models/
│   └── sample_list.txt

HO3D

Download and unzip the HO3D dataset following the official instructions to the data/HO3D folder. if everything is correct, the HO3D & YCB folder in your data should look like this:

data/
├── HO3D/
│   ├── evaluation/
│   ├── evaluation.txt
│   ├── train/
│   └── train.txt
├── YCB_models/
│   ├── 002_master_chef_can/
│   ├── ...

Download our YCB_models_supp and place it at data/YCB_models_supp

Now the data folder should have a root structure like:

data/
├── fhbhands/
├── fhbhands_supp/
├── fhbhands_example/
├── HO3D/
├── YCB_models/
├── YCB_models_supp/

5. Download pre-trained checkpoints

download our pre-trained CPF_checkpoints, unzip it at the CPF_checkpoints folder:

CPF_checkpoints/
├── honet/
│   ├── fhb/
│   ├── ho3dofficial/
│   └── ho3dv1/
├── picr/
│   ├── fhb/
│   ├── ho3dofficial/
│   └── ho3dv1/

6. Launch visualization

We create a FHBExample dataset in hocontact/hodatasets/fhb_example.py that only contains 10 samples to demonstrate our pipeline. Notice: this demo requires active screen for visualizing. Press q in the "runtime hand" window to start fitting.

$ python training/run_demo.py \
    --gpu 0 \
    --init_ckpt CPF_checkpoints/picr/fhb/checkpoint_200.pth.tar \
    --honet_mano_fhb_hand

7. Test on full dataset (FHB, HO3D v1/v2)

We provide shell srcipts to test on the full dataset to approximately reproduce our results.

FHB

dump the results of HoNet and PiCR:

$ python training/dumppicr_dist.py \
    --gpu 0,1 \
    --dist_master_addr localhost \
    --dist_master_port 12355 \
    --exp_keyword fhb \
    --train_datasets fhb \
    --train_splits train \
    --val_dataset fhb \
    --val_split test \
    --split_mode actions \
    --batch_size 8 \
    --dump_eval \
    --dump \
    --vertex_contact_thresh 0.8 \
    --filter_thresh 5.0 \
    --dump_prefix common/picr \
    --init_ckpt CPF_checkpoints/picr/fhb/checkpoint_200.pth.tar

and reload the GeO optimizer:

# setting 1: hand-only
$ CUDA_VISIBLE_DEVICES=0,1,2,3 python training/optimize.py \
    --n_workers 16 \
    --data_path common/picr/fhbhands/test_actions_mf1.0_rf0.25_fct5.0_ec \
    --mode hand

# setting 2: hand-obj
$ CUDA_VISIBLE_DEVICES=0,1,2,3 python training/optimize.py \
    --n_workers 16 \
    --data_path common/picr/fhbhands/test_actions_mf1.0_rf0.25_fct5.0_ec \
    --mode hand_obj \
    --compensate_tsl

HO3Dv1

dump:

$ python training/dumppicr_dist.py  \
    --gpu 0,1 \
    --dist_master_addr localhost \
    --dist_master_port 12356 \
    --exp_keyword ho3dv1 \
    --train_datasets ho3d \
    --train_splits train \
    --val_dataset ho3d \
    --val_split test \
    --split_mode objects \
    --batch_size 4 \
    --dump_eval \
    --dump \
    --vertex_contact_thresh 0.8 \
    --filter_thresh 5.0 \
    --dump_prefix common/picr_ho3dv1 \
    --init_ckpt CPF_checkpoints/picr/ho3dv1/checkpoint_300.pth.tar

and reload optimizer:

# hand-only
$ CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python training/optimize.py \
    --n_workers 24 \
    --data_path common/picr_ho3dv1/HO3D/test_objects_mf1_likev1_fct5.0_ec/ \
    --lr 1e-2 \
    --n_iter 500 \
    --hodata_no_use_cache \
    --lambda_contact_loss 10.0 \
    --lambda_repulsion_loss 4.0 \
    --repulsion_query 0.030 \
    --repulsion_threshold 0.080 \
    --mode hand

# hand-obj
$ CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python training/optimize.py \
    --n_workers 24 \
    --data_path common/picr_ho3dv1/HO3D/test_objects_mf1_likev1_fct5.0_ec/ \
    --lr 1e-2 \
    --n_iter 500  \
    --hodata_no_use_cache \
    --lambda_contact_loss 10.0 \
    --lambda_repulsion_loss 6.0 \
    --repulsion_query 0.030 \
    --repulsion_threshold 0.080 \
    --mode hand_obj

HO3Dofficial

dump:

$ python training/dumppicr_dist.py  \
    --gpu 0,1 \
    --dist_master_addr localhost \
    --dist_master_port 12356 \
    --exp_keyword ho3dofficial \
    --train_datasets ho3d \
    --train_splits val \
    --val_dataset ho3d \
    --val_split test \
    --split_mode official \
    --batch_size 4 \
    --dump_eval \
    --dump \
    --test_dump \
    --vertex_contact_thresh 0.8 \
    --filter_thresh 5.0 \
    --dump_prefix common/picr_ho3dofficial \
    --init_ckpt CPF_checkpoints/picr/ho3dofficial/checkpoint_300.pth.tar

and reload optimizer:

$ CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python training/optimize.py \
    --n_workers 24 \
    --data_path common/picr_ho3dofficial/HO3D/test_official_mf1_likev1_fct\(x\)_ec/  \
    --lr 1e-2 \
    --n_iter 500 \
    --hodata_no_use_cache \
    --lambda_contact_loss 10.0 \
    --lambda_repulsion_loss 2.0 \
    --repulsion_query 0.030 \
    --repulsion_threshold 0.080 \
    --mode hand_obj

Results

Testing on the full dataset may take a while ( 0.5 ~ 1.5 day ), thus we also provide our test results at fitting_res.txt.

K-MANO

We provide pytorch implementation of our Kinematic-chained MANO in lixiny/manopth, which is modified from the original hassony2/manopth. Thank Yana Hasson for providing the code.

Citation

If you find this work helpful, please consider citing us:

@article{yang2020cpf,
  title={CPF: Learning a Contact Potential Field to Model the Hand-object Interaction},
  author={Yang, Lixin and Zhan, Xinyu and Li, Kailin and Xu, Wenqiang and Li, Jiefeng and Lu, Cewu},
  journal={arXiv preprint arXiv:2012.00924},
  year={2020}
}

And if you have any question or suggestion, do not hesitate to contact me through siriusyang[at]sjtu[dot]edu[dot]cn.

Comments

FileNotFoundError: [Errno 2] No such file or directory: 'assets/mano/MANO_RIGHT.pkl'

I executed this command: python training/run_demo.py --gpu 0 --init_ckpt CPF_checkpoints/picr/fhb/checkpoint_200.pth.tar --honet_mano_fhb_hand

So, I moved assets/mano folder to the path CPF/manopth/mano/webuser/ But, I am still getting the error

opened by anjugopinath 3
AttributeError: 'ParsedRequirement' object has no attribute 'req'
Could you tell me which version of Anaconda to use please? I am getting the below error:

neptune:/s/red/a/nobackup/vision/anju/CPF$ conda env create -f environment.yaml Collecting package metadata (repodata.json): done Solving environment: done

==> WARNING: A newer version of conda exists. <== current version: 4.9.2 latest version: 4.10.1

Please update conda by running

$ conda update -n base -c defaults conda

Preparing transaction: done Verifying transaction: done Executing transaction: done Installing pip dependencies: | Ran pip subprocess with arguments: ['/s/chopin/a/grad/anju/.conda/envs/cpf/bin/python', '-m', 'pip', 'install', '-U', '-r', '/s/red/a/nobackup/vision/anju/CPF/condaenv.agtpjn0v.requirements.txt'] Pip subprocess output: Collecting git+https://github.com/utiasSTARS/liegroups.git (from -r /s/red/a/nobackup/vision/anju/CPF/condaenv.agtpjn0v.requirements.txt (line 1)) Cloning https://github.com/utiasSTARS/liegroups.git to /tmp/pip-req-build-ey_prxpa Obtaining file:///s/red/a/nobackup/vision/anju/CPF/manopth (from -r /s/red/a/nobackup/vision/anju/CPF/condaenv.agtpjn0v.requirements.txt (line 12)) Obtaining file:///s/red/a/nobackup/vision/anju/CPF (from -r /s/red/a/nobackup/vision/anju/CPF/condaenv.agtpjn0v.requirements.txt (line 13)) Collecting trimesh==3.8.10 Using cached trimesh-3.8.10-py3-none-any.whl (625 kB) Collecting open3d==0.10.0.0 Using cached open3d-0.10.0.0-cp38-cp38-manylinux1_x86_64.whl (4.7 MB) Collecting pyrender==0.1.43 Using cached pyrender-0.1.43-py3-none-any.whl (1.2 MB) Collecting scikit-learn==0.23.2 Using cached scikit_learn-0.23.2-cp38-cp38-manylinux1_x86_64.whl (6.8 MB) Collecting chumpy==0.69 Using cached chumpy-0.69.tar.gz (50 kB)

Pip subprocess error: Running command git clone -q https://github.com/utiasSTARS/liegroups.git /tmp/pip-req-build-ey_prxpa ERROR: Command errored out with exit status 1: command: /s/chopin/a/grad/anju/.conda/envs/cpf/bin/python -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-hnf78qhk/chumpy/setup.py'"'"'; file='"'"'/tmp/pip-install-hnf78qhk/chumpy/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(file);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-pip-egg-info-k7bp5gq7 cwd: /tmp/pip-install-hnf78qhk/chumpy/ Complete output (7 lines): Traceback (most recent call last): File "", line 1, in File "/tmp/pip-install-hnf78qhk/chumpy/setup.py", line 15, in install_requires = [str(ir.req) for ir in install_reqs] File "/tmp/pip-install-hnf78qhk/chumpy/setup.py", line 15, in install_requires = [str(ir.req) for ir in install_reqs] AttributeError: 'ParsedRequirement' object has no attribute 'req' ---------------------------------------- ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.

failed

CondaEnvException: Pip failed
opened by anjugopinath 3
How to use CPF on both hands?

Thanks a lot for your great work! I have a question: Since you only use the MANO_RIGHT.pkl, it seems that CPF currently can only construct right hand model, right? What is needed to be modified to use CPF on both hands? Thanks!

opened by buaacyw 3
Error when executing command "conda env create -f environment.yaml"
Hi,

I get the below error when executing the command "conda env create -f environment.yaml"

CondaError: Downloaded bytes did not match Content-Length url: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/linux-64/pytorch-1.6.0-py3.8_cuda10.2.89_cudnn7.6.5_0.tar.bz2 target_path: /home/anju/anaconda3/pkgs/pytorch-1.6.0-py3.8_cuda10.2.89_cudnn7.6.5_0.tar.bz2 Content-Length: 564734769 downloaded bytes: 221675180
opened by anjugopinath 1
Some questions about PiQR code

In the contacthead.py, the three decoders have different input dimension. self.vertex_contact_decoder = PointNetDecodeModule(self._concat_feat_dim, 1) self.contact_region_decoder = PointNetDecodeModule(self._concat_feat_dim + 1, self.n_region) self.anchor_elasti_decoder = PointNetDecodeModule(self._concat_feat_dim + 17, self.n_anchor)

I am wondering if this part is used to predict selected anchor points within each subregion.

The classification of subregions is obtained by contact_region_decoder and then the anchor points are predicted by anchor_elasti_decoder, is it right ?

I am a little bit confused about it, because according to the paper, Anchor Elasticity (AE) represents the elasticities of the attractive springs. But in the code, the output of anchor_elasti_decoder has no relation to the elasticity parameter, I'm wondering if there's some part I've missed.

Sorry for any trouble caused and thanks for your help!

opened by lym29 0
what's the meaning of "adapt"?

I notice that there are hand_pose_axisang_adapt_np and hand_pose_axisang_np in your code. Could you please explain what's the difference between them?

opened by Yamato-01 5
Expected code date ?

Hi !

I just read through your paper, congratulation on the great work ! I love the fact that you provide an anatomically-constrained MANO, and the per-object-vertex hand part affinity.

I look forward to the code realease :)

Do you have a planned date in mind ?

All the best,

Yana

opened by hassony2 4

Releases(v1.0.0)

v1.0.0(Dec 4, 2022)

This release is meant to archive the models, checkpoints, datasets, and results of CPF in ICCV 2021
Source code(tar.gz)
Source code(zip)

Owner

Lixin YANG

PhD student @ SJTU. Computer Vision, Robotic Vision and Hand-obj Interaction

GitHub Repository

Accompanying code for the paper "A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment".

#backdoor-HSIC (bd_HSIC) Accompanying code for the paper "A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment". To generate

0 Nov 25, 2021

Modeling CNN layers activity with Gaussian mixture model

GMM-CNN This code package implements the modeling of CNN layers activity with Gaussian mixture model and Inference Graphs visualization technique from

3 Aug 05, 2022

This is a model to classify Vietnamese sign language using Motion history image (MHI) algorithm and CNN.

Vietnamese sign lagnuage recognition using MHI and CNN This is a model to classify Vietnamese sign language using Motion history image (MHI) algorithm

3 Feb 24, 2022

A package, and script, to perform imaging transcriptomics on a neuroimaging scan.

Imaging Transcriptomics Imaging transcriptomics is a methodology that allows to identify patterns of correlation between gene expression and some prop

10 Dec 27, 2022

SCAAML is a deep learning framwork dedicated to side-channel attacks run on top of TensorFlow 2.x.

SCAAML (Side Channel Attacks Assisted with Machine Learning) is a deep learning framwork dedicated to side-channel attacks. It is written in python and run on top of TensorFlow 2.x.

69 Dec 21, 2022

SeqAttack: a framework for adversarial attacks on token classification models

A framework for adversarial attacks against token classification models

23 Nov 25, 2022

RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection

RODD Official Implementation of 2022 CVPRW Paper RODD: A Self-Supervised Approach for Robust Out-of-Distribution Detection Introduction: Recent studie

17 Oct 11, 2022

WeakVRD-Captioning - Implementation of paper Improving Image Captioning with Better Use of Caption

30 Oct 28, 2022

Vision-Language Transformer and Query Generation for Referring Segmentation (ICCV 2021)

Vision-Language Transformer and Query Generation for Referring Segmentation Please consider citing our paper in your publications if the project helps

143 Dec 23, 2022

Code of Puregaze: Purifying gaze feature for generalizable gaze estimation, AAAI 2022.

PureGaze: Purifying Gaze Feature for Generalizable Gaze Estimation Description Our work is accpeted by AAAI 2022. Picture: We propose a domain-general

39 Dec 05, 2022

[NeurIPS 2021] Shape from Blur: Recovering Textured 3D Shape and Motion of Fast Moving Objects

[NeurIPS 2021] Shape from Blur: Recovering Textured 3D Shape and Motion of Fast Moving Objects YouTube | arXiv Prerequisites Kaolin is available here:

107 Dec 26, 2022

Metric learning algorithms in Python

metric-learn: Metric Learning in Python metric-learn contains efficient Python implementations of several popular supervised and weakly-supervised met

1.3k Dec 28, 2022

4st place solution for the PBVS 2022 Multi-modal Aerial View Object Classification Challenge - Track 1 (SAR) at PBVS2022

A Two-Stage Shake-Shake Network for Long-tailed Recognition of SAR Aerial View Objects 4st place solution for the PBVS 2022 Multi-modal Aerial View Ob

5 Nov 09, 2022

LLVIP: A Visible-infrared Paired Dataset for Low-light Vision

LLVIP: A Visible-infrared Paired Dataset for Low-light Vision Project | Arxiv | Abstract It is very challenging for various visual tasks such as image

[email protected]"> 377 Jan 07, 2023

Pytorch implementation of “Recursive Non-Autoregressive Graph-to-Graph Transformer for Dependency Parsing with Iterative Refinement”

Graph-to-Graph Transformers Self-attention models, such as Transformer, have been hugely successful in a wide range of natural language processing (NL

40 Aug 14, 2022

A PyTorch implementation of Radio Transformer Networks from the paper "An Introduction to Deep Learning for the Physical Layer".

An Introduction to Deep Learning for the Physical Layer An usable PyTorch implementation of the noisy autoencoder infrastructure in the paper "An Intr

120 Nov 21, 2022