I am a Senior Research Scientist  (Level A) at the Jet  Propulsion Laboratory/NASA, California Institute of Technology, located in  Pasadena, CA, USA. I also hold an appointment as an Associate Professor at the Univ. of Southern California (webpage is located here ). I have a PhD from the Dept.  of Electrical and Computer Engineering, Carnegie Mellon University where I  worked with Prof. David Casasent.  I obtained my M.S. at Iowa State University where I worked with Dr. Jennifer  Davidson . I am a principal investigator in several  projects funded by DARPA, NASA, Dept. of Homeland Security, U.S. Army, commercial organizations, and the NIH, in a  variety of areas spanning multidisciplinary areas in computer science and engineering, ranging from robotics, biometrics, intelligent sensor networks, sensor fusion, data mining, and system optimization and control to machine learning, computer vision, pattern recognition, and image  processing. I have directly supervised, and managed over 16 researchers comprised of research senior staff and post-docs at USC and JPL including recruiting and building the team, conducting administrative oversight, technical leadership, and providing vision and guidance to the entire team. 

I have more than 65 publications in journals and conference  proceedings. I have given 2 plenary lectures at international conferences and workshops, a colloquial lecture at the University of South Florida, several invited talks at conferences and academic institutions, and 3 Keynote Addresses at conferences. I am a member of the technical organizing committee of the Optical  Pattern Recognition Conference in the Annual SPIE Defense and Security Symposium  (formerly known as the Annual SPIE  AeroSense Conference), held in Orlando, FL every year during  April. I have chaired several conference sessions (including sessions on  Optical Pattern Recognition and " Active  vision in robotics " at Photonics East,  Intelligent Robotics and Computer Vision XVII), and am a reviewer on several  journals (IEEE Trans. Image Processing, IEEE Transactions on Signal Processing,  IEEE Transactions on Systems, Man, and Cybernetics-B, Applied Optics,  Neurocomputing, Neural Networks, and Optical Engineering).  My detailed research  accomplishments are listed in my CV (not up to date!!).                                                                                         .  

Research Interests and Activities

My interests are in the general broad area of Autonomous Systems, including:

• Robotics and unmanned systems with autonomy
• Multimedia and Cross-Media Data mining, Search and Retrieval
• Standards and evaluation for multimedia data analysis, search and retrieval (contact me for our science datasets used for evaluation of cross-media data analysis for event detection and tracking)
• Image, signal, and video processing
• Wireless Sensor Networks, Wireless sensors and systems for space, health, science, and defense 
• Autonomy in Remote Wireless Systems and Embedded Systems
• Control optimization, Scheduling and Resource Management Theory, Real-time distributed control, resource allocation with low latency
• Machine learning , Active Learning, Transfer Learning, Artificial Intelligence
• Machine vision, image processing , signal and multidimensional processing
• Neural networks, Pattern recognition, Statistics, Stochastic modeling, support vector machines and kernel machines
• Feature extraction and dimensionality reduction from structured and unstructured data
• Sensor fusion and real-time multisensor processing and adaptation
• Numerical, mathematical, and algebraic methods applied to diverse fields
• Active learning
• Novel embedded signal processing for low power sensors and instruments
• Embedded control and processing combined with novel hardware designs 
• Algorithms and algorithm development

Applications and fields I've worked in over the past several years include:

• Sensor networks for mobile health monitoring, environmental monitoring, and defense
• Biometrics and surveillance systems
• Multi-band adaptive software defined radio hardware design with small (mobile) form-factors for extremely low-power applications
• Distributed optimal control of multiple wireless heterogeneous sensor nodes
• Adaptive multiple software defined radio control and resource allocation
• Robotics, active vision/perception, and active learning
• Human-machine interaction, face detection and recognition
• Image processing applied to various fields, speech processing and recognition
• Biomedical image and signal processing
• Materials Science
• Sensors and sensor fusion
• Automatic target recognition
• Product inspection

My PhD dissertation at Carnegie Mellon University entitled " Nonlinear Feature Extraction for Pattern Recognition Applications" involved the design of new machine learning techniques for nonlinear kernel-based feature extraction that I developed independently of support vector machines and kernel-PCA, and have shown to outperform SVMs and kernel PCA on highly nonlinear, high-dimensional data sets such as images and multi-dimensional data. You can download a copy of my PhD thesis here. If you need a PDF file, please email me at Ashit.Talukder@jpl.nasa.gov and I will gladly send you a PDF file of my thesis.

I've led and worked on several industrial and federal funded projects at JPL, USC, and CMU . A few of the relevant projects are listed below. While this list below is not comprehensive, it covers some of areas that I have worked on in the past.

Biometrics and Surveillance Systems

We have built state of the art biometrics and surveillance systems for several federal sponsors. The system is a multimodal long-standoff biometrics system, that can detect human at a long standoff distance, identify them using the multimodal measurements. We have unique capability to detect a human behind a wall and also determine the identity of an individual through walls at a long distance (compared to existing solutions that can only detect the presence of a person behind walls).  Due to the sensitive nature of this work, the technical details of the biometrics and surveillance system cannot be provided online.

Wireless Intelligent Sensor Networks and Software Defined Radios with real-time Resource Allocation Capabilities

In this task, we have developed a wireless adaptive sensor network hardware testbed with nodes that run for days, developed distributed sensing algorithms using novel pattern recognition techniques and sensor fusion algorithms, and designed and implemented new control optimization and resource allocation algorithms for real-time distributed end-to-end system management. Our primary contributions are:

• A multi-node sensor network testbed with generic COTS sensor hookup capability where each node can last for days or weeks (see the hardware components and sensors currently in place in our sensor network in Figure 2 below)
• A novel heirarchical sensor network architecture (see Figure 1 below)where higher levels have increased functional and computational capabilities
• Real-time event detection and data prioritization capability at every node using pattern recognition techniques
• Sensor fusion  from heterogeneous distributed sensors at higher levels in the sensor network, thereby enabling robust event detection and intelligent decision making
• Novel Genetic algorithm and Model-predictive control based control optimization solutions for integrated resource management in our sensor network.   
• Visualization tools at Level 3 that automatically mark detected events as alarms. This will enable experts to focus in on specific events and time-intervals of interest. See Figure 3 below.

                                                     Wireless distributed sensor network architecture

    Levels 1 and 2                             Levels 1, 2,  and 3 in the sensor network

             Event visualization tools in Level 3 of our sensor network

Adaptive Control and Resource Management in Multiple Distributed Sensor Networks and Software Defined Radios 

We have designed and implemented novel control algorithms for real-time adaptive operation of multiple RF nodes in dynamic  environments where the channel noise and ambient RF noise characteristics can change dramatically based on the relative postion and motion between nodes. We use novel Model Predictive Control Algorithms on RF nodes with reasonable storage/memory resources, and Markov Decision Processes on RF nodes with extremely limited storage/limited resources (less than 4K of memory) for adaptive resource management and coordinated operation of the SDR nodes, where each node has different computing, momory and RF characteristics. We believe our work in this area to be the first demonstrable example of real-time adaptive control of distributed wireless nodes and software-programmable radios with such limited computing and storage resources .

Coordination of Unmanned Mobile Robots with Wireless Sensor Networks for Coastal Environmental Monitoring  

We developed a novel framework for controlling operational parameters of both fixed sensors and mobile unmanned robots  in real-time and using the resulting sensor observations to increase the predictive accuracy of pre-computed mathematical models of the ocean environment. The sensor controller uses the Model Predictive Control technique to focus limited sensor resources to the regions that are most likely to impact the accuracy of the model forecasts. The system was demonstrated on an offline simulation of the environment of the New York harbor and ocean. The sensor network adaptation and resource management concept for coastal monitoring networks are shown below.

Demonstration of adaptive control of 4 underwater unmanned vehicles (UUVs) (red and blue dots) in response 
to a flooding event (green regions) as it evolves over time

Cross-Media Streaming Data Mining for Cyclone Detection and Tracking Using Knowledge Transfer and Machine Learning  

Tropical and extra-tropical cyclones are important components of the Earth climate system that exhibit variability at different temporal and spatial scales. A cyclone landfall causes great devastation, incurs fatality, and affects people's livelihood. To identify and track tropical weather system, the Tropical Prediction Center/National Hurricane Center (TPC/NHC) uses conventional surface and upper-air observations and reconnaissance aircraft reports (Pasch, Stewart, and Brown 2003), and these are concentrated in the North American coasts and in Japan/Europe to some degree. Coverage on a global basis, especially in under-developed and developing nations such as large portions of Asia and Africa is limited or lacking which results in disastrous consequences in many of these regions. In recent years, some studies have used satellite images that are manually retrieved and analyzed to improve the accuracy of cyclone tracking; this procedure is currently slow, tedious, involves coverage of only local regions in North America, and requires close analysis by teams of experts.  

Mining of remote satellite data poses to be a truly challenging task, like most multimedia data mining problems. The datasets have different resolutions, varying dimensionality (1-D, 2-D or higher), and are often irregularly sampled (see examples below)

However, QuikSCAT single sensor measurements have some drawbacks. Due to the polar orbiting nature of the QuikSCAT satellite, it has limited spatial and temporal coverage. In particular, the time interval between two consecutive observations of a cyclone is approximately around 12 hours. Sometimes the sensor can only capture partial measurement of the cyclone due to the orbiting path. To alleviate these problems, one can use sensor measurements from multiple orbiting satellites. However, sensors from other satellites measure other earth and atmospheric properties. These measurements are not as powerful in identifying cyclone as the wind properties measured by the Scatterometer on QuikSCAT satellite. For example, the precipitation rate measured by the Tropical Rain Measurement Mission (TRMM)[1] has been used to trace the cyclone track manually either during a cyclone event or as a reanalysis after a cyclone event. However, the precipitation rate cannot be used for cyclone identification since heavy rainfall does not imply a cyclone event. Moreover, since precipitation rate is not a definite cyclone indicator, there is no archive of positively labeled (cyclone) examples from the TRMM data by the scientific community. the main contribution is a methodology for transferring local spatial-temporal knowledge between satellite data from different data sources to enable event detection using weaker features. To avoid negative transfer for our methodology, the assumption that the weaker feature, which is not a definite discriminating feature, has to be able to pick up the event of interest must be satisfied. In other words, without applying our methodology, using the weaker feature for object detection/identification should result in a high false positive rate and high true positive rate, i.e. low precision but high recall. This methodology also significantly improves the temporal resolution for cyclones to 3 hr accuracies and resolves a large number of event occlusion instances that would occur in single sensor QuikSCAT-based cyclone tracking.

Concept for Knowledge transfer learning between multiple satellites for more accurate cyclone detection & tracking

Autonomous cyclone detection and tracking for Hurricane Gonu in the Arabian Sea in 2005 using knowledge sharing 
between QuikSCAT and TRMM. Red square is the final detected cyclone - Green and black squares correspond to rejected false alarms.

Mobile Continuous Adaptive Decision Making and Learning for Autonomous Sensor Networks

We have implemented novel pattern recognition algorithms for real-time continuous adaptive monitoring of dynamic environments from distributed heterogeneous RF nodes and sensors, and demonstrated this on real-time mobile telehealth monitoring. While several publications exist on this work in a variety of technical conferences and workshops, the most relevant media coverage of the applicability of our system in a real-world domain can be found in the CrossBow 2005 Second Quarter Industrial Newsletter at  http://www.xbow.com/General_info/Info_pdf_files/XbowNewsletter_Q2-05.pdf

Dynamic Scene Perception from Mobile Robots in Urban Environments

I am a Co-Investigator in competed & funded DARPA IPTO Robotics 2020 contract Object-Referenced Robot Navigation in Dynamic Urban Environments (Total Award: $4.0 million). Collaborating institutions: JPL, Carnegie Mellon University, SRI, UC Santa Cruz. Start date: Sept. 2002 : End: Sept 2004. Project involves detection of moving objects in real-time from mobile robot platforms using vision sensors, and localization of robots in urban terrain using high-level percept referenced navigation commands specified by user.

Robot Navigation and Control in Natural Terrain

DARPA funded MARS (Mobile Autonomous Robot Software) project involving design of algorithms to be used for autonomous vehicles and robots for use in urban environments and natural terrain. Tasks include autonomous terrain classification and obstacle detection using 3-D modeling and vision, computer graphics, geometrical reasoning, and pattern recognition. Details of our ongoing DARPA MARS work at JPL can be found here. My work also ties in with the DARPA ATO Tactical Mobile Robotics (TMR) program at JPL that involves the JPL Urban Robot (URBIE), DARPA PERCEPTOR., and DEMO III.

Face Detection and Facial Feature Extraction

Face detection and facial feature location funded by NGMTec and Graphco Technologies for synthetic generation of video based on audio input for use in low-bandwidth videoconferencing, and speech-to-video applications

Biomedical Applications

• Biomedical image processing algorithms and software for plaque analysis and segmentation of teeth to analyze and visulize plaque removal efficacy for measuring the effectiveness of new toothbrush designs prior to mass-manufacture. This software is expected to dramatically speed up design, testing, and time-to-manufacture of new toothbrushes, thereby significantly reducing design, development, and testing costs.
• Data analysis and processing of images of cells for life detection, and planetary protection to swiftly evaluate and compare techniques for "cleaning" spacecraft prior to launch to ensure that other plantary systems are not seeded with live earth organisms.
• Real-time eyetracking algorithm and system development (NASA/Congress funded) using IR sources, IR filters, and FPGA implementation to assist handicapped people to use computers, evaluate alertness of astronauts on the space station, and diagnose health defects such as Amblyopia

Automatic Target Recognition, Industrial Automation and Robotics

• DARPA ATD/R program for robust algorithm and sensor fusion for automatic target recognition (ATR) from FLIR imagery using wavelets, morphology, and correlation filters
• NSF-funded MIMP (Mesoscale Interface Mapping Project) at MRSEC (Materials Research Science and Engineering Center) at CMU for real-time processing of SEM microscopy images of mesoscale structures to predict future failure in materials. Collaborative work with CMU Material Sciences Dept., Mathematics Dept., and ECE Dept.
• USDA funded product inspection system design and development for non-destructive inspection of agricultural products from X-ray images using new image object detection, image segmentation, and image classification with advanced neural network approaches.
• New nonlinear feature extraction technique implemented as a neural network for feature extraction and classification of multi-dimensional data. Theoretically and experimentally proven to yield better nonlinear transforms than neural networks or kernel-machines on high-dimensional data sets.

 Selected Publications

• S.-S. Ho and A. Talukder, Automated Cyclone Discovery and Tracking using Multiple Heterogeneous Satellite Data, 14th ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, Las Vegas, NV, 24-27 Aug, 2008.
• S.-S. Ho and A. Talukder, Cyclone Tracking Using Multiple Satellite Data Sources via Spatial-Temporal Knowledge Transfer, AAAI-08 workshop, Transfer Learning for Complex Tasks, Chicago, IL, 14 July, 2008. 
• A. Talukder, S.-S. Ho, T. Liu, W. Tang, A. Bingham, and E. Rigor, Global Cyclone Detection and Tracking using Multiple Remote Satellite Data, NASA Earth Science Technology Conference, Aldephi, MD, 24 June, 2008. 
• S.-S. Ho and A. Talukder, Automated Cyclone Identification from Remote QuikSCAT Satellite Data, IEEE Aerospace Conference 2008, Big Sky, MT, Mar. 1-8, 2008.
• A. Talukder,  S. Ali and A. Panangadan, "Markov Decision Processes for Control of a Sensor Network-based Health Monitoring System", Accepted for presentation as an Emerging Application paper at the Seventeenth Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-05).
• A. Talukder, S. Monacos, T. Sheikh, Distributed multisensor processing and classification under constrained resources for mobile health monitoring and remote environmental monitoring, Proceedings of International Conference on Intelligent Sensing and Information Processing, January 2004.
• A. Talukder, R. Bhatt, T. Sheikh, R. Pidva, L. Chandramouli, and S. Monacos, Dynamic control and power management algorithm for continuous wireless monitoring in sensor networks, In Proceedings of the 29th Conference on Local Computer Networks, Workshop on Embedded Network Sensors (EmNetS-I), pp. 498-505, 2004.
• A. Talukder, R. Bhatt, L. Chandramouli, T. Sheikh, R. Pidva, and S. Monacos, Autonomous Resource Management and Control Algorithms for Distributed Wireless Sensor Networks, The 3^rd ACS/IEEE International Conference on Computer Systems and Applications - January 2005
• “Optimal sensor scheduling and power management in sensor networks”, A. Talukder, SPIE Defense and Security Symposium, Optical Pattern Recognition XI, April 2005.
• Featured Article on CrossBow 2005 Second Quarter Industrial Newsletter at  http://www.xbow.com/General_info/Info_pdf_files/XbowNewsletter_Q2-05.pdf
• “Interleaving Wavelet Coefficients for Reliable Data Reconstruction in Low-Bandwidth Sensor Networks”, Anand Panangadan, Ashit Talukder, SenSys 2005 (Submitted).
• “Eye-tracking architecture for biometrics and remote monitoring”, Ashit Talukder, John-Michael Morookian, Steve Monacos, Raymond Lam, Clayton LeBaw, James L. Lambert, Applied Optics, Vol 44, No. 5, pp. 693-700, Feb 2005.
• “Obstacle Detection and Terrain Classification for Autonomous Off-Road Navigation”, R. Manduchi, A. Castano, A. Talukder, L. Matthies, Autonomous Robots, 18:81-102, 2005
• “Real-time visual odometry and moving object segmentation from moving robots", A. Talukder, L. Matthies, IROS 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems
• A. Talukder and D. Casasent, “ A closed-form neural network for discriminatory feature extraction from high-dimensional data ”, Neural Networks , Vol. 14 (9), Pages 1201-1218, November 2001.
• “ General Methodology for Simultaneous Representation and Discrimination of Multiple Object Classes”, A. Talukder and D. Casasent, Optical Engineering (Advances in Recognition Techniques) , Mar. 1998.
• A. Talukder, S. Goldberg, L. Matthies, and A. Ansar, “Real-time detection of moving objects in a dynamic scene from moving robotic vehicles”, Accepted for Oral Presentation at the IEEE Intelligent Robots and Systems Conference (IEEE IROS) 2003.
• A. Talukder, R. Manduchi, A. Rankin, L. Matthies, " Fast and Reliable Obstacle Detection and Segmentation for Cross-country Navigation ", IEEE Intelligent Vehicles Conference 2002, France, June 18-22, 2002.
• A. Talukder, R. Manduchi*, R. Castano, K. Owens, L. Matthies, A. Castano, R. Hogg, " Autonomous Terrain Characterisation and Modelling for Dynamic Control of Unmanned Vehicles", IEEE IROS 2002, Switzerland, Sept 30-Oct 2, 2002.
• “ Nonlinear features for robotics, inspection, and face recognition ”, (Keynote Address) D. Casasent and A. Talukder, Proc. SPIE , vol. 3804, Jul. 1999.
• “ Detection and Segmentation of Items in X-Ray Imagery ”, David Casasent, A. Talukder, P. Keagy, T. Schatzki. Accepted for publication in Transactions ASAE , 2000.
• " Classification of pistachio nuts from X-ray images", Casasent, Talukder, Schatzki and Keagy, Accepted (with modifications) at Lebensmittel-Wissenschaft und-Technologie (Intl. Journal for Food, Science and Tehcnology), November 1999.
• " Adaptive Activation Function Neural Net for Face Recognition", Ashit Talukder, David Casasent, IEEE Intl Joint Conf. on Neural Networks , pp. 549-552. 2001.
• “ Pose Invariant Recognition of Faces with Unknown Pose ”, A. Talukder and D. Casasent, Intl. Joint Conf. Neural Networks (IJCNN) 1999 (and journal paper in preparation), Jul. 1999.
• “Distortion-Invariant Object Representation and Discrimination Using an FST Neural Net”, D. Casasent, M. Sipe and A. Talukder, 1998 Intl. Joint Conf. on Neural Networks (IJCNN’98), May 1998.
• " Neural Net with Adaptive Activation Functions for Face Recognition ", Ashit Talukder, David Casasent, IEEE Intl Joint Conf. on Neural Networks , 2000. 
• "Accurate Multiscale Gabor Wavelet Fusion for Edge Detection in Microscopy Images" (Invited Paper), A. Talukder, D. P. Casasent, Proc. SPIE, Wavelet Applications V, 3391, Apr. 1998 (Also printed in Selected Key SPIE Papers on CD-ROM series, Vol 8: Mathematical Imaging and Vision , Ed. Dr. Gerhard Ritter, Dec. 1999).
• "Automated Estimation of Class and Pose of Machined Parts", A. Talukder and D. Casasent, Robotics and Machine Perception Newsletter , 1999.
• "Pose Estimation and Transformation of Faces from Single Views", A. Talukder and D. Casasent, Proc. SPIE: Robots and Computer Vision XVII , Nov. 1998
• " Classification and Pose Estimation of Objects using Nonlinear Features ", A. Talukder and D. Casasent, Proc. SPIE: Applications and Science of Computational Intelligence , Vol. 3390, Apr. 1998.
• " Classification of Product Inspection Items Using Nonlinear Features ", A. Talukder and D. Casasent, Proc. SPIE, Optical Pattern Recognition IX , Vol. 3386, Apr. 1998.
• " X-Ray Sensor Agricultural Product Inspection" (Invited Article), A. Talukder and D. Casasent. Robotics and Machine Perception Newsletter (special issue on Machine Vision), Vol. 7, No. 1, 1998, p. 9-11.
• " Modified Binary Watershed Transform for Segmentation of Agricultural Products ", Ashit Talukder, David Casasent, Ha-Woon Lee, Pamela M. Keagy and Thomas F. Schatzky, Proc. SPIE , Nov 1998.
• "X-Ray Agricultural Product Inspection: Segmentation and Classification", D. Casasent, A. Talukder, H.W. Lee. Proc. SPIE, Intelligent Systems & Advanced Manufacturing , 3205, Oct. 1997.
• " Image processing for grain boundary detection in microscope images ", A. Talukder, D. Casasent and S. Ozdemir, Proc. Intl. Grain Growth Conf. (ICGG-3) , Jun. 1998.
• "Detection of bands in backscatter microscopy images using new Hough transform techniques", D. Casasent, L. Chen and A. Talukder, Proc. SPIE, Hybrid Image and Signal Proc. VI , Vol. 3389, Apr. 1998.
• "Real-Time Robust Line Detection in Microscopy Images", A. Talukder and D. Casasent. Proc. SPIE (Intelligent Systems and Advanced Manufacturing), 3208, Oct. 1997.
• "Joint Recognition and Discrimination in Nonlinear Feature Space", A. Talukder and D. Casasent. Proc. SPIE (Intelligent Systems and Advanced Manufacturing) , 3208, Oct. 1997.
• "Automated Segmentation and Feature Extraction of Product Inspection Items", A. Talukder and D. Casasent. Proc. SPIE (AeroSense), Apr. 1996.
• "Algorithm fusion for detection with reduced false alarms", Casasent D., Ye A., Talukder A. Proc. SPIE (Optical Pattern Recognition VII), vol.2752, p. 206-213
• "Feature space trajectory neural net classifier: confidences and thresholds for clutter and low contrast objects", Neiberg L., Casasent D., Talukder A. Proc. SPIE (Applications and Science of Artificial Neural Networks II) , vol.2760, p. 435-46.
• "Detection algorithm fusion concepts for computer vision", Casasent D., Anqi Ye, Talukder, A. Proc. SPIE (Intelligent Robots and Computer Vision XIV), vol.2588, p. 2-9.
• "Model selection and texture segmentation using partially ordered Markov models", Talukder, A., Davidson, J. 1995 International Conference on Acoustics, Speech, and Signal Processing (ICASSP-95), p. 2527-2530.
• "Texture analysis using partially ordered Markov models", Davidson, J., Talukder, A., Cressie, N. Proc. ICIP-94 (Proceedings of 1st International Conference on Image Processing), 1994, p. 402-406.

• Receipient of NASA Space Act Award 2003 and monetary award for software and algorithm development for Intra-plaque analysis system.
• Provisional Patent (NPO-30417) “Automated Intra-oral plaque analysis system” filed through the California Institute of Technology Intellectual Property Office for work on Colgate-Palmolive funded PlaqTrak System (Nov 2001).
• Provisional Patent (JPL and NASA Case No. NPO 30699) “ New Real-Time Slice-based Processing and Eye-Gaze Mapping Software Application“, filed through the California Institute of Technology Intellectual Property Office for work on NASA-funded Eyetracker for Biomedical Applications.
• Provisional Patent (JPL and NASA Case No. NPO 30700) “ Novel System Software and Hardware Architecture for Optimized Real-time Non-invasive Eyetracking“, filed through the California Institute of Technology Intellectual Property Office for work on NASA-funded Eyetracker for Biomedical Applications.
• NASA NTR # 40377, “Real-time optical flow computation for computer vision-based applications” filed through the California Institute of Technology Intellectual Property Office for work on DARPA-funded Dynamic Scene Perception project
• Premium Award for Academic Excellence (PACE) received at Iowa State University

Ashit's Resume (CV)

If you have any questions, or want to contact me, please send me email at: Ashit.Talukder@jpl.nasa.gov