Natsuko "Suko" Toyofuku
I have been in the Klein lab since I entered the Vision Science PhD program in 2000. I am working on a project involving various models for frequency detection and classification images. I am concurrently working in Dr. Cohn's Visual Detection Lab on another project where I examine the cues to detect object looming/expansion. I also work at the US Department of Agriculture, where I work with x-ray images.
Current Klein Lab project (part of my thesis):
Frequency Detection Model Testing With Classification Images For Suprathreshold StimuliKeywords: image analysis, model testing, classification images, frequency detection, multiple channel model, structured noise
Classification images, an auditory analysis technique adapted to Vision Science experiments by Ahumada in 1996, allow scientists to capture what a subject is picturing in their minds, or at least examine which aspects of a visual stimulus the subject uses most often to make a detection decision. These aspects are not always a perfect match to the stimulus and often represent a strategy of conserving attentional resources to those aspects which most reliably give correct answers.
This technique relies on adding a noise mask (usually random white noise) to the stimulus and examining the images that led to errors. There are often tens of thousands of trials. The noise pixels which were confused with the stimulus, can be identified by this technique.
This project focused on using a variation on this technique to test three models for frequency detection. We modified the technique to use far fewer trials and with a structured (not random) noise mask. The stimuli were four combinations of a 1 cy/deg and 3 cy/deg grating pattern. The noise was the 1st-7th harmonics of a base 0.5 cy/deg fundamental.
The models we were testing are the single channel mode, a max rule multiple channel model (from Graham and Nachmias 1971) and our proposal, a matched filter multiple channel model.
We also measured contrast thresholds, efficiency, consistency, dprime, and power, to further test the model predictions.
Education:
- ScB in Cognitive Science from Brown University, Providence, RI
- Advanced to PhD Candidacy (I hope to file my dissertation in Fall 2005), Vision Science, UC Berkeley, Berkeley, CA
Research Interests:
- Human perception, particularly signal detection and pattern recognition in noise.
- X-ray image analysis
- Detection strategies
Research Background:
Physical Scientist, US Dept of Agriculture, Agricultural Research Service (7+ years and for the last 5 years I have worked there concurrently with studies at UCB)
With the USDA, I am working on image-based quality control measures in fruits, wheat and meat. I primarily work with x-ray images although my lab has worked with NIR, spectroscopy and color sorting.
My current project with the USDA is examining xray images of international passenger luggage for agricultural contraband (such as meat and fresh fruit). We are hoping to develop a list of image based features (rather than object based features) that will assist in luggage inspection. By image-based, we mean looking for round shapes in the image rather than looking for what the inspectors think an x-rayed apple or orange looks like. Objects look very different in xray, because of lack of color, parallax and large amounts of compression and we are not used to seeing the insides of objects. We would like this list of features to be used to develop a faster and more homogenous training technique than is in place currently, and perhaps become the beginnings of an automated computer pattern detection program to pre-sort luggage images to assist the inspectors.
Ph.D Graduate Student, Dr. Stanley A. Klein, Vision Science Program (5 years, see blurb above)
Ph.D Graduate Student, Dr. Theodore E. Cohn, Vision Science Program (5 years, concurrent with work in Klein lab)
With Dr. Cohn, I am working on human detection of looming/expanding objects. The impetus behind this project is to help reduce rear end collisions with buses (traffic collision avoidance scenarios). To assist in the design of a rear-mounted braking signal, we tested the sensitivity to expansion of edges and textures in the horizontal and vertical dimensions.
Publications, Abstracts, Conference Presentations:
PAPERS and REPORTS
Spanjer, M. C., Scholten, J. M., Kastrup, S., Jörissen, U., Schatzki, T. F. and Toyofuku, N.(in press) "Sample Comminution for Mycotoxin Analysis: Dry Milling or Slurry Mixing?" Journal of Food Additives and Contaminants, accepted for publication, May 2005.
Schatzki, T., Toyofuku, N. (2004) "Sampling And Sample Preparation Of Pistachios." Book Chapter in Meeting The Mycotoxin Menace. Ed : Barug, D., Van Egmond, H.P., Lopez-Garcia, R., Van Osenbruggen, W.A., Visconti, A. Netherlands: Wageningen Academic Publishers. Pp.. 221-235.
Schatzki, T.F. and Toyofuku, N. (2003) "Sampling and Sample Preparation of Pistachios." Proceedings of the Second World Mycotoxin Forum. Waageningen, Holland.
Schatzki, T.F. and Toyofuku, N. (2003) "Sample Preparation and Pre-Sampling of Pistachios." Journal Of Agriculture And Food Chemistry. 51(20):6068-6072.
Toyofuku, N. and Nguyen, T. (2002) "Did You See That? Monocular Factors in Perception of Looming Objects," current literature review submitted for grant quarterly report, May 2002.
Schatzki, T.F. and Toyofuku, N. (2002) "Subsample Preparation of Pistachios." Proceedings of the 6th International Fruit, Nut and Vegetable Production Engineering Symposium. Potsdam, Germany. Pp. 171-176.
Toyofuku, N., Klein, S.A. and Carney, T. (2001) "First Plus Third Harmonic Summation in Noise", Journal of Vision, 1(3), abstract no. 445.
Haff, R.P. and Toyofuku, N. (2000) "Economical Method for X-ray Inspection of Wheat", ASAE technical presentation, Milwaukee, WI.
Schatzki, T.F. and Toyofuku, N. (2000) "Subsampling in Pistachios", Crop Management and Handling, Insect Control and Fungal Relationships, Session 2: 66-67.
Pearson, T. and Toyofuku, N. (2000) "Automated Sorting of Pistachio Nuts with Closed Shell", Applied Engineering in Agriculture, 16 (1): 91-94.
Toyofuku, N. and Schatzki, T.F. (1999) "Method for Batch Testing Red Tart Cherries for the Presence of Pit Fragments", Journal of Food Processing and Preservation, 23 (5): 435-441.
Schatzki, T. F., Haff, R. P., Young, R., Can, I., Le, L.-C. and Toyofuku, N. (1997) "Defect Detection in Apples by Means of X-ray Imaging", Transactions of the ASAE, 40 (5): 1407-1415.
Schatzki, T. F., Haff, R. P., Young, R., Can, I., Le, L.-C. and Toyofuku, N. (1997) "Defect Detection in Apples by Means of X-ray Imaging", SPIE Proc., 2907.
Presentations and Posters
Toyofuku, N. and Schatzki, T.F. (2005) "Feasibility of Feature-Based Contraband Detection in X-Ray Images," poster presented at Vision Sciences Society Conference, Sarasota, FL.
Toyofuku, N., Klein, S.A. and Carney, T. (2004) "Templates are in the Eye of the Beholder," poster presented at Vision Sciences Society Conference, Sarasota, FL.
Toyofuku, N. and Cohn, T.E. (2003) "Increased Sensitivity to Looming Objects With the Addition of Texture," poster presented at International Conference honoring David Regan at the Center for Vision Research (CVR) in Toronto, Canada.
Toyofuku, N. and Klein, S. A. (2002) "Internal and External Noise Contributions to Classification Templates: A Double Pass Analysis," poster presented at the Fall Vision Meeting, San Francisco, CA.
Toyofuku, N., Nguyen, T., and Cohn, T. E. (2002) "Transient Size Change Detection," poster presented at Vision Sciences Society Conference, Sarasota, FL.
Nguyen, K. N., Toyofuku, N., Nguyen, T. and Cohn, T. E. (Jan, 2002) "A Study and Utilization of Monocular Cues to Time to Collision," talk presentation at Driver Behavior Meeting, Berkeley, CA.
Toyofuku, N., Nguyen, K. N., Nguyen, T. and Cohn, T. E. (Oct, 2001) "Human Perceptual Task in Automated-Manual Longitudinal Control Transition", talk presentation at PATH Program-Wide Research Meeting, Davis, CA.
Toyofuku, N., Klein, S.A. and Carney, T. (2001) "First Plus Third Harmonic Summation in Noise", poster presented at Vision Sciences Society Conference, Sarasota, FL.
Email: suko@berkeley.edu