The Artificial Perception Laboratory

40 St. George Street, Room 4158 - Phone: (416)-946-8813
APL Infrastructures
UofT Explorer
Interactive Visual Search
FPGA-Based Speech Processing
VLSI Accelerated Sound Localization
Distributed Signal Processing for Sensor Networks
Probabilistic Speech Enhancement
Phase-Based Speech Processing
Sound Localization Using Distributed Microphone Arrays
Acoustically Guided Robots

The Edward S. Rogers Sr. Department of Electrical and Computer Engineering
University of Toronto
Artificial Perception Laboratory Projects








UofT Explorer
The UofT Explorer project utilizes an algorithm designed to quickly rank 2-dimensional images according to the user’s preference. 65 images are taken around the UofT St. George campus at different angles and elevations. When the user clicks on a point on any image, images that contain the selected area are given priority in the display. The algorithm depends on the proximity of labeled buildings to localize the area where the user selects. Images are ranked according to how well they show the area chosen by the user.




Interactive Visual Search
This project fuses information from a large array of cameras or database of images in order to create a three-dimensional unified representation of all objects captured by the images.

Representative Publications:
1. Mavandadi, S., Aarabi, P., Multi-Sensor Information Fusion with Application to Multi-Camera Systems, Proceedings of the 2004 IEEE International Conference on Systems, Man and Cybernetics (SMC 2004), The Hague, Netherlands, October 2004.
2. Aarabi, P., Localization-Based Sensor Validation Using The Kullback-Leibler Divergence. IEEE Transactions on Systems, Man, and Cybernetics Part B, Vol. 34, No. 2, pp. 1007-1016, April 2004. "pdf" "ps.gz"




FPGA-Based Speech Processing
A real-time hardware optimized implementation of speech localization and enhancement is performed using an Altera Stratix FPGA.

Representative Publications:
1. Halupka, D., Rabi, S. A., Aarabi, P., Sheikholeslami, A., Real-Time Dual-Microphone Speech Enhancement Using Field Programmable Gate Arrays. Proceedings of the 2005 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2005), Philadelphia, March 2005. "pdf"
2. Nguyen, D., Aarabi, P., Sheikholeslami, A., Real-Time Sound Localization Using Field-Programmable Gate Arrays. Proceedings of the 2003 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2003), Hong Kong, April 2003. "pdf" "ps.gz"




VLSI Accelerated Sound Localization
A custom implementation of the Phase Transform sound localization algorithm is developed using a 0.18um CMOS process. The implementation allows for real-time sound localization and source tracking on handheld computers and tablet PCs.

Representative Publications:
1. Halupka, D., Mathai, N., Aarabi, P., Sheikholeslami, A., Robust Sound Localization in 0.18um CMOS, IEEE Transactions on Signal Processing, Vol. 53, No. 6, June 2005. "pdf"
2. Aarabi, P., Mavandadi, S., Multi-Source Time Delays of Arrival Estimation Using Conditional Time-Frequency Histograms, Information Fusion, Vol. 4, No. 2, pp. 111-122, June 2003.



Distributed Signal Processing for Sensor Networks
This project involves the development and testing of a wireless network of sensors for distributed signal processing. Using the Crossbow Inc. Motes sensors, a novel algorithm and implementation is performed which focuses on real-time distributed spectrum estimation and time-delay estimation.

Representative Publications:
1. Jahromi, O., Aarabi, P., Theory and Design of Multirate Sensor Arrays, IEEE Transactions on Signal Processing, Vol. 53, No. 5, May 2005. "pdf"
2. Jahromi, O., Aarabi, P., Distributed Spectrum Estimation In Sensor Networks. Proceedings of the 2004 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2004), Montreal, May 2004. "pdf" "ps.gz"



Probabilistic Speech Enhancement
In this project, variational learning and probabilistic graphical models are employed to perform microphone-array based speech enhancement.

Representative Publications:
1. Rennie, S., Achan, K., Frey, B., Aarabi, P., Variational Speech Separation of More Sources than Mixtures. Proceedings of the Tenth International Workshop on Artificial Intelligence and Statistics (AISTATS'05), Barbados, January 2005.
2. Rennie, S., Aarabi, P., Kristjansson, T., Frey, B., Achan, K., Robust Variational Speech Separation Using Fewer Microphones Than Speakers. Proceedings of the 2003 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2003), Hong Kong, April 2003. "pdf" "ps.gz"



Phase-based Speech Processing
Most speech recognition and enhancement systems utilize only the magnitude of the recorded speech signals. This project involves utilizing only the phase of the recorded signals for speech recognition and enhancement.

Representative Publications:
1. Aarabi, P., Shi, G. Phase-Based Dual-Microphone Robust Speech Enhancement, IEEE Transactions on Systems, Man, and Cybernetics Part B, Vol. 34, No. 4, pp. 1763-1773, August 2004. "pdf" "ps.gz"
2. Lai, C., Aarabi, P., Multiple-Microphone Time-Varying Filters For Robust Speech Recognition. Proceedings of the 2004 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2004), Montreal, May 2004. "pdf" "ps.gz"
3. Shi, G., Aarabi, P., Robust Digit Recognition Using Phase-Dependent Time-Frequency Masking. Proceedings of the 2003 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2003), Hong Kong, April 2003. "pdf" "ps.gz"




Sound Localization Using Distributed Microphone Arrays
A distributed array of 48 microphones is setup for the real-time localization and tracking of speakers in the Artificial Perception Lab.

Representative Publications:
1. Mungamuru, B., Aarabi, P., Enhanced Sound Localization, IEEE Transactions on Systems, Man, and Cybernetics Part B, Vol. 34, No. 3, pp. 1526-1540, June 2004. "pdf" "ps.gz"
2. Aarabi, P., The Fusion of Distributed Microphone Arrays for Sound Localization, EURASIP Journal of Applied Signal Processing (Special Issue on Sensor Networks), Vol. 2003, No. 4, pp. 338-347, March 2003. "pdf" "ps.gz"
3. Aarabi, P., Self-Localizing Dynamic Microphone Arrays, IEEE Transactions on Systems, Man, and Cybernetics Part C, Vol. 32, No. 4, pp. 474-484, November 2002. "pdf" "ps.gz"




Acoustically Guided Robots
This project involves utilization of microphone arrays and speech processing for guiding and interfacing with mobile robots.

Representative Publications:
1. Wang, Q.H., Ivanov, T., Aarabi, P., Acoustic Robot Navigation Using Distributed Microphone Arrays. Information Fusion (Special Issue on Robust Speech Processing), Vol. 5, No. 2, pp. 131-140, June 2004.
2. Aarabi, P., Wang, Q.H., Yeganegi, M., Integrated Displacement Tracking And Sound Localization. Proceedings of the 2004 IEEE Conference on Acoustics, Speech, and Signal Processing (ICASSP 2004), Montreal, May 2004. "pdf" "ps.gz"



Link to Fingerprint Recognition project.
Link to old Projects page.

© Copyright 2005 the Artificial Perception Laboratory