Redwood Neuroscience Institute

If you haven't heard of it yet, check it out. They have made available a ton of their previous seminar videos and a great deal of them deal with vision as it is central to one of their goals. Some highlights include Dana Ballard, Jeff Hawkins, and Thomas Serre.

Trevor Darrell - Visual Recognition and Tracking for Perceptive Interfaces

http://www.researchchannel.org/prog/displayevent.aspx?rID=6939&fID=345

Devices should be perceptive, and respond directly to their human user and/or environment. In this talk I'll present new computer vision algorithms for fast recognition, indexing, and tracking that make this possible, enabling multimodal interfaces which respond to users' conversational gesture and body language, robots which recognize common object categories, and mobile devices which can search using visual cues of specific objects of interest. As time permits, I'll describe recent advances in real-time human pose tracking for multimodal interfaces, including new methods which exploit fast computation of approximate likelihood with a pose-sensitive image embedding. I'll also present our linear-time approximate correspondence kernel, the Pyramid Match, and its use for image indexing and object recognition, and discovery of object categories. Throughout the talk, I'll show interface examples including grounded multimodal conversation as well as mobile image-based information retrieval applications based on these techniques.

Erik Sudderth - Learning Hierarchical, Nonparametric Models for Visual Scenes

http://www.researchchannel.org/prog/displayevent.aspx?rID=24390&fID=345

Computer vision systems use image features to detect and categorize objects in visual scenes. In this University of Washington program, learn about Erik Sudderth MIT/UC Berkeley research that explores hierarchical models using contextual and geometric relationships for more effective learning from large, partially labeled image databases.

Tony Jebara - From Perception and Discriminative Learning to Interactive Behavior

http://www.researchchannel.org/prog/displayevent.aspx?rID=2748&fID=345

A strong symbiosis lies between machine learning and machine perception. Just as we learn to reason and interact with the world through our senses, a smart sensing system could acquire data to drive higher level learning problems. Ironically, learning and probabilistic methods themselves can provide the driving machinery for perception as well. I demonstrate several examples of probabilistic sensors in wearable and room-based environments. These human-centered systems perform object detection, face tracking, 3d modeling, recognition, and topic-spotting in real-time.

Aude Oliva - Understanding Visual Scenes in 200 msec: Results from Human and Modeling Experiments

http://www.researchchannel.org/prog/displayevent.aspx?rID=5953&fID=345

One of the remarkable aspects of human image understanding is that we are able to recognize the meaning of a novel image very quickly and independently of the complexity of the image. This talk will review findings in human perception that help us understand which mechanisms the human brain uses to achieve fast visual recognition, accurate visual search and adequate memorization of visual information. It also will describe the limits of human perception, as well as how to use our understanding of the pros and cons of these mechanisms for designing artificial vision systems and visual displays for human use.

Paul Bloom - Conscious of the Present; Conscious of the Past: Language (cont.); Vision and Memory

This lecture finishes the discussion of language by briefly reviewing two additional topics: communication systems in non-human primates and other animals, and the relationship between language and thought. The majority of this lecture is then spent on introducing students to major theories and discoveries in the fields of perception, attention and memory. Topics include why we see certain visual illusions, why we don't always see everything we think we see, and the relationship between different types of memory.

Valentina Anna Maria Daelli - Short-term memory effects on visual perception

http://videolectures.net/eccs08_daelli_stmeovp/

Memory traces, stored in the form of attractors or appearing as the result of recent perceptual experience, can actively shape processing and categorization of visual stimuli. Electrophysiology in monkey IT cortex and computational modeling indicate the existence of categorical boundaries in the dynamics of cortical networks. Our psychophysical experiments in humans show that these boundaries can be shifted following recent visual experience, in adaptation and priming paradigms.

Malcolm Slanley - Hallucinations in Auditory Perception

http://videolectures.net/mbc07_slaney_hap/

In this talk I want to review the need for richer architectures for auditory processing. Many experiments point to the tangled web of connections in the perceptual system, yet our engineering solutions remain almost exclusively bottom-up. How is it that we can provide context, so that our systems can solve musical analysis and auditory scene-analysis problems? I'll talk about notable systems that are successful "hallucinators."

Kenji Doya - Computational Models of Basal Ganglia Function

As a mathematical engineer, Kenji Doya approaches the goal of describing the most intricate brain mechanisms from a computational perspective. He constructs models of reinforcement learning involving the networked structures of the basal ganglia. His efforts are captured and expressed quantitatively as probabilities, regressions, and algorithms.

In this presentation, Doya covers basic concepts of reinforcement learning, then surveys the last decade of inquiry into the components of the basal ganglia circuit governing voluntary motion. Among the topics: action values, action candidates, and reward prediction involving the neurotransmitter dopamine; model-free versus model-based learning strategies; and the essential role of serotonin as modulator in the complex information loop.

Doya’s recent research is carried out via robots he calls “cyber rodents.” His dream as an undergraduate was to “build a robot that learns the variety of behaviors on its own.” That is, the computer, not the human engineer, teaches the robot to move. He accomplished this in designing a machine-creature exhibiting emotion-like attributes characterized as “depression,” “impulsivity,” “greed,” and “patience.”

Doya believes the “metaparameters” of reinforcement learning must be “tuned appropriately…Otherwise the performance of your learning is very, very poor.” The iterative process involves three terms -- the reward itself, the expected reward for a new state based on choice of action, and memory of the reward gained in the previous state. In the comparison, any differential greater than zero can be exploited for learning. The tradeoff: “No pain, no gain.”

As research advanced to increasing levels of structural specificity, Doya posited that “there seems to be spatial segregation in the function” of basal ganglia components. Specialization in aspects of reinforcement learning is now seen, for instance, in ventral versus dorsal areas of the striatum.

Differentiation is also found in the cortico-basal ganglia information network: not a simple closed loop, but parallel electrical pathways conducting distinct neural operations. Further, the neuromodulators each have their respective missions. Dopamine encodes the temporal difference error -- the reward learning signal. Acetylcholine affects learning rate through memory updates of actions and rewards. Noradrenaline controls width or randomness of exploration. Serotonin is implicated in “temporal discounting,” evaluating if a given action is worth the expected reward. Doya reminds us that clinically “it is well known that the serotonin function is impaired in the depression patient.”

The system of basal ganglia components and neuromodulators requires dynamic balancing. A delicate interplay determines outcomes for learning, actions, and affective states. Doya’s synthetic models are proxies for human behavior, and his computational framework describing the moving parts ultimately has therapeutic implications for psychiatric and neurological disorders.

Takao K Hensch - Critical Period Mechanisms of Visual Cortical Plasticity

http://videocast.nih.gov/launch.asp?13781

Abstract: (CIT): With interests ranging from child development and human disease to brain circuitry, Dr. Hensch is best known for his studies of the interplay of sensory input and genetics early in mammalian brain development, when the unusual malleability of the nervous system allows experiences to shape the lifelong wiring of the brain. His research has employed techniques from systems and molecular neuroscience to probe the mechanisms of early neural wiring, the limits of early brain plasticity, and how such plasticity might be restored later in life. Such work could have profound implications for developmental disorders as well as learning and education. Dr. Hensch's group made the surprising finding that the maturation of inhibitory circuits - such as the stunted development of wiring in visual cortex from an eye deprived of vision - controls the timing of early brain plasticity. By directly manipulating the onset of such inhibitory transmission within the brain, Dr. Hensch and his colleagues have shown that neural plasticity mirrors this timing, a finding that has had a major impact on developmental neuroscience. He continues to investigate the structural and molecular mechanisms of these phenomena, as well as the role of rhythmic electrical activity that occurs in the brain during sleep, visual experience, and brain plasticity. NIH Neuroscience Seminar Series.

Steve Buck & Scott Murray - Vision and the Brain

Why do we need vision? As it turns out, there are two answers to this question. On the one hand, we need vision to give us detailed knowledge of the world beyond ourselves, knowledge that allows us to recognize things from minute to minute and day to day. On the other hand, we also need vision to guide our actions in that world at the very moment they occur. These are two quite different job descriptions, and nature seems to have given us two different visual systems to carry them out. Dr. Murray explores the problems in the context of object size and brightness perception, and discusses computational challenges in sight that require extensive neural processing.

http://www.researchchannel.org/prog/displayevent.aspx?rid=16249

Oliver Sacks - Seeing Hallucinations while Blind

Pawan Sinha - Project Prakash

http://web.mit.edu/bcs/sinha/prakash.html



‘Prakash’ in Sanskrit means light.
The goal of Project Prakash is to bring light into the lives of curably blind children and, in so doing, illuminate some of the most fundamental scientific questions about how the brain develops and learns to see.

Nanci Bell - The Role of Imagery and Verbal Processing

Nanci Bell, creator of a program which aims to stimulate gestalt imagery in order to aid in language comprehension and analytical thinking, discusses the role imagery plays in comprehension in individuals of all ages. Series: "M.I.N.D. Institute Lecture Series on Neurodevelopmental Disorders"