Learning Go/NoGo Terrain Classification
نویسنده
چکیده
This paper presents a software system for imagebased terrain classification that mimics a human supervisor’s segmentation and classification of training images into “Go” and “NoGo” regions. The system identifies a set of image chips in the training images that span the range of terrain appearance. It then uses these exemplars to segment novel images and assign fuzzy Go/NoGo classification. System parameters adapt to new inputs, providing a mechanism for learning. Index Terms – terrain classification, computer vision, machine learning, exemplar memory
منابع مشابه
P300 Component Modulation During a Go/Nogo Task in Healthy Children
ABSTRACT Introduction: Several differences in the P300 component are observed when responses must be executed or inhibited in the Go/Nogo task. However, few studies were established by using well-controlled task with respect to the preparatory processing and stimulus probability. In the present study, we examined the peak amplitude and latency of Go-P300 (P300 evoked by visual Go stimuli) and N...
متن کاملIntact associative learning in patients with schizophrenia: evidence from a Go/NoGo paradigm.
OBJECTIVE Schizophrenia is associated with deficits in executive control and associative learning. In the present study, we investigated the effect of associative learning during a Go/NoGo task in healthy controls subjects and patients with schizophrenia. METHODS Thirty patients with schizophrenia and 30 age-and-gender matched healthy control subjects performed 15 blocks of training and 3 blo...
متن کاملPii: S1388-2457(98)00036-4
Objective: To study cortical potentials associated with suppression of intended motoric actions. Methods: Electro-encephalographic activity was recorded in a Go/NoGo reaction time paradigm. Subjects viewed computer-generated pacing stimuli, which provided information concerning the time at which an imperative Go/NoGo signal occurred. A motoric response was required following Go stimuli while mo...
متن کاملPavlovian influences on learning differ between rats and mice in a counter-balanced Go/NoGo judgement bias task
Judgement bias tests of animal affect and hence welfare assume that the animal's responses to ambiguous stimuli, which may herald positive or negative outcomes, are under instrumental control and reflect 'optimism' or 'pessimism' about what will happen. However, Pavlovian control favours responses (e.g. approach or withdrawal) according to the valence associated with a stimulus, rather than the...
متن کاملThe Effects of Amnestic Mild Cognitive Impairment on Go/NoGo Semantic Categorization Task Performance and Event-Related Potentials.
We examined the effects of amnestic mild cognitive impairment (aMCI) on behavioral (response times and error rates) and scalp-recorded event-related potential (ERP) measures of response execution and inhibition, using Go/NoGo tasks involving basic and superordinate semantic categorization. Twenty-five aMCI (16 F; 68.5±8 years) and 25 age- and gender-matched normal control subjects (16 F; 65.4±7...
متن کامل