https://github.com/baophann/phone_advance

https://github.com/baophann/phone_advance

Science Score: 26.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
  • Academic publication links
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (6.0%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

Basic Info
  • Host: GitHub
  • Owner: BaophanN
  • License: mit
  • Language: Python
  • Default Branch: main
  • Size: 44.9 KB
Statistics
  • Stars: 0
  • Watchers: 0
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created 10 months ago · Last pushed 10 months ago
Metadata Files
Readme License

README.md

Quantized Distillation for Driver Activity Recognition

This is the official PyTorch implementation of our IROS 2023 paper:

Quantized Distillation: Optimizing Driver Activity Recognition Models for Resource-Constrained Environments

Setup

Add folder called model in the same directory as above code to save trained checkpoints. Before training, the folder structure should look like this

├── [path to your cloned repository] ├── architecture ├── config ├── framework_activity_recognition ├── model # add this ├── LICENSE ├── README.md └── requirement.txt Do not forget to install the requirement stated in the folder.

Dataset

Drive&Act

Pretrained Model

MobileNet3D | RGB I3D

Training

To train baseline RGB I3D model on Drive&Act, use the following command python -m framework_activity_recognition config/train/i3dbaseline.yaml To train other baseline or using knowledge distillation and/or quantization on Drive&Act, replace the yaml file in the command to one of the following yaml file in config/train folder

├── ./config ├── /train ├── i3dbaseline.yaml # RGB I3D baseline on Drive&Act ├── mobilenet_quant.yaml # MobileNet3D with PyTorch quantization on Drive&Act ├── mobilenetbaseline.yaml # MobileNet3D baseline on Drive&Act ├── studentteacher.yaml # MobileNet3D on Drive&Act with knowledge distillation from teacher RGB I3D └── studentteacher_quant.yaml # MobileNet3D with PyTorch quantization and knowledge distillation from teacher RGB I3D on Drive&Act

Test

To test RGB I3D Model with test split of Drive&Act, use the following command python -m framework_activity_recognition config/test/i3dtest.yaml To test another model, replace the yaml file in the command with one of the following yaml file in config/test folder

├── ./config ├── /test ├── i3dtest.yaml # RGB I3D test on Drive&Act test split ├── mobilenetquanttest.yaml # MobileNet3D with PyTorch quantization test on Drive&Act test split └── mobilenettest.yaml # MobileNet3D test on Drive&Act test split

Owner

  • Name: Bao Phan
  • Login: BaophanN
  • Kind: user

Where there's a will, there's a way.

GitHub Events

Total
  • Push event: 4
  • Create event: 2
Last Year
  • Push event: 4
  • Create event: 2

Dependencies

requirement.txt pypi
  • Markdown ==3.3.6
  • Pillow ==8.4.0
  • PyYAML ==6.0
  • Werkzeug ==2.0.3
  • absl-py ==1.0.0
  • cachetools ==4.2.4
  • certifi ==2021.10.8
  • charset-normalizer ==2.0.12
  • dataclasses ==0.6
  • google-auth ==2.6.0
  • google-auth-oauthlib ==0.4.6
  • grpcio ==1.43.0
  • idna ==3.3
  • imageio ==2.15.0
  • importlib-metadata ==4.8.3
  • numpy ==1.19.5
  • oauthlib ==3.2.0
  • opencv-python ==4.5.5.62
  • pandas ==1.1.5
  • protobuf ==3.19.4
  • pyasn1 ==0.4.8
  • pyasn1-modules ==0.2.8
  • python-dateutil ==2.8.2
  • pytz ==2021.3
  • requests ==2.27.1
  • requests-oauthlib ==1.3.1
  • rsa ==4.8
  • scipy ==1.5.4
  • six ==1.16.0
  • sk-video ==1.1.10
  • tensorboard ==2.8.0
  • tensorboard-data-server ==0.6.1
  • tensorboard-plugin-wit ==1.8.1
  • torch ==1.10.2
  • torchaudio ==0.10.2
  • torchvision ==0.11.3
  • typing_extensions ==4.1.0
  • urllib3 ==1.26.8
  • zipp ==3.6.0