Research

I am interested in computer vision, deep learning, generative AI, and virtual worlds. My most recent research is on modeling virtual humans using cameras. Here is a summary of my research:

	UPose3D: uncertainty-aware 3D human pose estimation with cross-view and temporal cues Vandad Davoodnia, Saeed Ghorbani, Marc-André Carbonneau, Alexandre Messier, Ali Etemad ECCV, 2024 We present an uncertainty-aware 3D pose estimation method through maximizing the likelihood of 3D keypoints based on the differentiable distributions of multi-view 2D keypoints. [Paper] [Project]
	SkelFormer: markerless 3D pose and shape estimation using skeletal transformers Vandad Davoodnia, Saeed Ghorbani, Alexandre Messier, Ali Etemad ECCV Workshop, 2024 We present SkelFormer, an inverse-kinematic regression framework for estimating human pose and shape from videos. [Paper] [Project]
	FabriCar: enriching the user experience of in-car media interactions with ubiquitous vehicle interiors using e-textile sensors Pouya M Khorsandi, Lee Jones, Vandad Davoodnia, Timothy J Lampen, Aliya Conrad, Ali Etemad, Sara Nabil ACM Designing Interactive Systems Conference (DIS), 2023 We investigate the use of e-textiles in cars to reduce driver distraction when performing non-driving tasks like media playback. [Paper] [PDF]
	Human pose estimation from ambiguous pressure recordings with spatio-temporal masked transformers Vandad Davoodnia, Ali Etemad IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2023 We propose a temporal variation of the ViTs for pose estimation and pre-train it using a masked auto-encoder framework. [Paper]
	Deep multitask learning for pervasive BMI estimation and identity recognition in smart beds Vandad Davoodnia, Monet Slinowsky, Ali Etemad Journal of Ambient Intelligence and Humanized Computing, 2020 We use neural networks for BMI estimation and user identification from in-bed pressure sensors. [Paper]
	Parse: Pairwise alignment of representations in semi-supervised eeg learning for emotion recognition Guangyi Zhang, Vandad Davoodnia, Ali Etemad IEEE Transactions on Affective Computing, 2022 We present PARSE for understanding emotions from EEG data when limited labeled data is available. [Paper] [Code]
	Estimating pose from pressure data for smart beds with deep image-based pose estimators Vandad Davoodnia, Saeed Ghorbani, Ali Etemad Journal of Applied Intelligence, 2021 We present a domain adaptation method to improve body pose detection from pressure data. [Paper]
	In-bed pressure-based pose estimation using image space representation learning Vandad Davoodnia, Saeed Ghorbani, Ali Etemad IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021 We address the problems of using standard pose estimation models on pressure maps by transforming unclear pressure maps into a format resembling human figures on normal images. [Paper]
	Computer-aided diagnosis using class-weighted deep neural network Pritam Sarkar, Vandad Davoodnia, Ali Etemad IEEE International Conference On Machine Learning And Applications (ICMLA), 2019 We present a deep learning method for diagnosing breast cancer tumors from medical images. [Paper]
	A novel recursive network for irony detection in tweets Arthur Cruz de Araujo Vandad Davoodnia, Guangyi Zhang, Ali Etemad, Xiaodan Zhu Course Report, 2019 We present a recursive attention-based Bi-LSTM neural network for irony detection in tweets. [Paper] [Code]
	Classification of hand movements from EEG using a deep attention-based LSTM network Guangyi Zhang, Vandad Davoodnia, Alireza Sepas-Moghaddam, Yaoxue Zhang, Ali Etemad IEEE Sensors Journal, 2020 We present a left and right hand movements classification method from EEG signals using LSTM network with added attention mechanism. [Paper]
	Identity and posture recognition in smart beds with deep multitask learning Vandad Davoodnia, Ali Etemad IEEE International Conference on Systems, Man and Cybernetics (SMC), 2019 We introduce a deep learning model that uses pressure sensor arrays to accurately identify a person and their sleep posture. [Paper]
	Prognostication of lignocellulosic biomass pyrolysis behavior using ANFIS model tuned by PSO algorithm Mortaza Aghbashlo, Meisam Tabatabaei, Mohammad Hossein Nadian, Vandad Davoodnia, Salman Soltanian Fuel, 2019 We developed an Adaptive Network-based Fuzzy Inference System (ANFIS) model to estimate biomass pyrolysis behavior and trained it using Particle Swarm Optimization (PSO). [Paper]
	Multifractal detrended fluctuation analysis of continuous neural time series in primate visual cortex Zahra Fayyaz, Mohammadreza Bahadorian, Jafar Doostmohammadi, Vandad Davoodnia, Sajad Khodadadian, Reza Lashgari Journal of Neuroscience Methods, 2019 We demonstrated a new method for response tuning of neural activity time-series. [Paper]
	Effect of age and glaucoma on the detection of darks and lights Linxi Zhao, Caroline Sendek, Vandad Davoodnia, Reza Lashgari, Mitchell W Dul, Qasim Zaidi, Jose-Manuel Alonso Investigative Ophthalmology & Visual Science (IOVS), 2015 We developed an iOS and Android App to study the effect of age and glaucoma on response time to light and dark targets. [Paper] [iOS] [Android]