121,830 research outputs found
ZONING DESIGN FOR HANDWRITTEN NUMERAL RECOGNITION
No full textMicrosoft, Motorola, Siemens, Hitachi, IAPR, NICI, IUF
In the field of Optical Character Recognition (OCR), zoning is used to extract topological information from patterns. In this paper zoning is considered as the result of an optimisation problem and a new technique is presented for automatic zoning. More precisely, local analysis of feature distribution based on Shannon's entropy estimation is performed to determine "core" zones of patterns. An iterative regiongrowing procedure is applied on the "core" zones to determine the final zoning
New Steps Toward the Effective Evaluation of e-learning Activities: A Participant-Based Approach
No full text
Growth performance and fatty acid metabolism in European grayling (Thymallus thymallus, L.) fed diets differing in n-3 and n-6 PUFA levels
No full textA feeding trial was carried out to evaluate the growth performance
and fatty acid metabolism of juvenile grayling (T. thymallus,
L., Adriatic strain) in response to three diets supplying varying
levels and ratios of essential fatty acids (n-3 PUFAs, 7, 12, 8 g/kg;
n-6 PUFAs, 5, 10, 12 g/kg; n-3/n-6 ratios, 1.6, 1.2, 0.6). Each diet
was fed to visual satiety over 9 weeks to triplicate fish groups,
kept at 13.5±0.4°C, each consisting of 25 specimens (average
wgt. 12±0.3 g). Regardless of the dietary n-3 PUFA content, fish
growth and feed efficiency improved (P<0.05) by increasing
dietary n-6 PUFA level or decreasing n-3/n-6 ratio. Based on in
Generation of Ensambles of Synthetic Classifiers for the Evaluation of Combination Methods
No full textThis paper presents a new technique for generating sets of synthetic classifiers to evaluate abstract-level combination methods. The sets differ in terms of both recognition rates of the individual classifiers and degree of similarity. For this purpose, each abstract-level classifier is considered as a random variable producing one class label as the output for an input pattern. From the initial set of classifiers, new slightly different sets are generated by applying specific operators, which are defined at the purpose. Finally, the sets of synthetic classifiers have been used to estimate the performance of combination methods for abstract-level classifiers. The experimental results demonstrate the effectiveness of the proposed approach
Affective states recognition through touch dynamics
No full textThis work exploits Touch Dynamics to recognize affective states of a user while using a mobile device. To the aim, the acquired touch pattern is segmented in swipes, successively a wide set of handcrafted features is computed to characterize the swipe. The affective analysis is obtained through machine learning techniques. Data have been collected developing a specific App designed to acquire common unlock Android touch patterns. In this way the user interaction has been preserved as the more natural and neutral possible in real environments. Affective state labels have been obtained adopting a well-known psychological questionnaire. Three affective states have been considered: anxiety, stress and depression. Tests, performed on 115 users, reported an overall accuracy of 73.6% thus demonstrating the viability of the proposed approach
A Controlled Benchmark of Video Violence Detection Techniques
No full textThis benchmarking study aims to examine and discuss the current state-of-the-art techniques for in-video violence detection, and also provide benchmarking results as a reference for the future accuracy baseline of violence detection systems. In this paper, the authors review 11 techniques for in-video violence detection. They re-implement five carefully chosen state-of-the-art techniques over three dierent and publicly available violence datasets, using several classifiers, all in the same conditions. The main contribution of this work is to compare feature-based violence detection techniques and modern deep-learning techniques, such as Inception V3
Sit-to-Stand Test for Neurodegenerative Diseases Video Classification
No full textIn this extended version of this paper, an automatic video diagnosis system for dementia classification is presented. Starting from video recordings of patients and control subjects, performing sit-to-stand test, the designed system is capable of extracting relevant patterns for binary discern patients with dementia from healthy subjects. The original system achieved an accuracy 0.808 by using the rigorous inter-patient separation scheme especially suited for medical purposes. This separation scheme provides the use of some people for training and others, different, people for testing. The implementation of features from the kinematic theory of rapid human movement and its sigma-lognormal model together with classic features increased the overall accuracy of the system to 0.947 F1 score. In addition, multi-class classification was performed with the aim of classifying neurodegenerative disease severities. This work is an original and pioneering work on sit-to-stand video classification for neurodegenerative diseases, its novelties are on phases segmentation, experimental setup and the application of kinematic theory of rapid human movements to sit-to-stand videos for neurodegenerative disease assessment
Cognitive visual tracking and camera control
No full textAbstract not availableNicola Bellotto, Ben Benfold, Hanno Harland, Hans-Hellmut Nagel, Nicola Pirlo, Ian Reid, Eric Sommerlade, Chuan Zha
- …
