Skip to content

Latest commit

 

History

History
133 lines (98 loc) · 4.1 KB

README.md

File metadata and controls

133 lines (98 loc) · 4.1 KB
Raw

TAO: Context Detection from Daily Activity Patterns using Temporal Analysis and Ontology

[paper (IMWUT 2023)] [talk (IMWUT 2023)] [demo video]

Authors: [Sudershan Boovaraghavan] [Prasoon Patidar] [Yuvraj Agarwal]

Abstract: Translating fine-grained activity detection (e.g., phone ring, talking interspersed with silence and walking) into semantically meaningful and richer contextual information (e.g., on a phone call for 20 minutes while exercising) is essential towards enabling a range of healthcare and human-computer interaction applications. Prior work has proposed building ontologies or temporal analysis of activity patterns with limited success in capturing complex real-world context patterns. We present TAO, a hybrid system that leverages OWL-based ontologies and temporal clustering approaches to detect high-level contexts from human activities. TAO can characterize sequential activities that happen one after the other and activities that are interleaved or occur in parallel to detect a richer set of contexts more accurately than prior work. We evaluate TAO on real-world activity datasets (Casas and Extrasensory) and show that our system achieves, on average, 87% and 80% accuracy for context detection, respectively. We deploy and evaluate TAO in a real-world setting with eight participants using our system for three hours each, demonstrating TAO’s ability to capture semantically meaningful contexts in the real world. Finally, to showcase the usefulness of contexts, we prototype wellness applications that assess productivity and stress and show that the wellness metrics calculated using contexts provided by TAO are much closer to the ground truth (on average within 1.1%), as compared to the baseline approach (on average within 30%).

Reference

Download Paper Here

BibTeX Reference:

@article{10.1145/3610896,
author = {Boovaraghavan, Sudershan and Patidar, Prasoon and Agarwal, Yuvraj},
title = {TAO: Context Detection from Daily Activity Patterns Using Temporal Analysis and Ontology},
year = {2023},
issue_date = {September 2023},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
volume = {7},
number = {3},
url = {https://doi.org/10.1145/3610896},
doi = {10.1145/3610896},
journal = {Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.},
month = {sep},
articleno = {87},
numpages = {32},
keywords = {ontology, deep Learning, activity recognition, Behavioral context recognition}
}

Installation:

1. Clone (or Fork!) this repository

git clone [email protected]:synergylabs/TAO.git

2. Create a virtual environment and install python packages

We recommend using conda. Tested on Ubuntu 20.04, with python 3.7.

conda create -n "tao" python=3.7
conda activate tao
conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=10.2 -c pytorch

python -m pip install -r requirements.txt

Usage:

1. Using activity log files:

Save the activity log files in the folder: data_collection_pipeline/real_gt_data/

cd ontological-Pipeline/

streamlit run run_onto_pipeline.py

2. Using REST API:

streamlit run API_server.py

Example usage:

import requests
import json
import time
import numpy as np
import sys

port = 8080
request_url = '0.0.0.0'

print("Testing service on port:", port)
st_times = int(time.time())
timestamps = np.arange(st_times,st_times+80,2)

activities =[]

responses = []
for i in range(40):
    response = requests.get(
        f"http://{request_url}:{port}/predict?timestamp={timestamps[i]}&activities={activities[i]}")
    responses.append(response.text)
response_dict = json.loads(response.text)

3. Link to pre-trained models.

https://tao-context-sensing.s3.us-east-2.amazonaws.com/artifacts.zip

Contact:

For more information please contact [email protected]