Welcome to the Pelli lab
How does the brain recognize objects, like letters, words, and faces? We've shown that the first step is independent detection of simple features by channels (PDF). But how do we put those features together? We've discovered that crowding is the failure of this binding, when too many features are included (HTML). This insight has enabled fast progress on old problems: Faces are recognized like words (HTML); Crowding determines reading rate (PDF). We are currently using a new feature-counting technique to analyze the feature-integration process underlying object recognition and reading. Our approach is mostly psychophysical. We measure thresholds on human observers. We are gearing up to do more brain imaging of the object recognition process. The pace is fast. Our experiments are simple. Our software is flexible. More often than not, we answer the question in a week. Many of these questions arise at the weekly lab meeting, where everyone participates in a free-for-all discussion of each person's project. We welcome visitors. Would you like to join us for a lab meeting?
Updated