Image-to-MIDI mapping based on dynamic fuzzy color segmentation for visually impaired people

Chen, Chieh‐Li; Liao, Yan-Fa; Tai, Chung-Li

doi:10.1016/j.patrec.2010.11.019

Cited by 9 publications

(8 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Apart from smart devices, the applications span to developing an OpenKWS system, content‐based search in conversations, sung‐word recognition [21] and audio database indexing [22]. With the development of IoT, researchers are using wireless sensors to analyse environmental sounds particularly for bird species recognition [23], obtain obstacle information for visually‐impaired people [24], audio surveillance [25], whale sound categorisation [26] and automatic snore detection [27].…”

Section: Related Workmentioning

confidence: 99%

Audio classification using braided convolutional neural networks

Sinha

Awasthi

Ajmera

2020

IET signal process.

View full text Add to dashboard Cite

Convolutional neural networks (CNNs) work surprisingly well and have helped drastically enhance the state-of-the-art techniques in the domain of image classification. The unprecedented success motivated the application of CNNs to the domain of auditory data. Recent publications suggest hidden Markov models and deep neural networks for audio classification. This study aims to achieve audio classification by representing audio as spectrogram images and then use a CNN-based architecture for classification. This study presents an innovative strategy for a CNN-based neural architecture that learns a sparse representation imitating the receptive neurons in the primary auditory cortex in mammals. The feasibility of the proposed CNN-based neural architecture is assessed for audio classification tasks on standard benchmark datasets such as Google Speech Commands datasets (GSCv1 and GSCv2) and the UrbanSound8K dataset (US8K). The proposed CNN architecture, referred to as braided convolutional neural network, achieves 97.15, 95 and 91.9% average recognition accuracy on GSCv1, GSCv2 and US8 K datasets, respectively, outperforming other deep learning architectures.

show abstract

Section: Related Workmentioning

confidence: 99%

Audio classification using braided convolutional neural networks

Sinha

Awasthi

Ajmera

2020

IET signal process.

View full text Add to dashboard Cite

show abstract

“…Several prior traversable area segmentation techniques have been developed around the assumption that the area immediately in front of the robot is initially traversable and thus they identify the pathway by comparison to a safe window near the bottom of the image [8,9,14]. This assumption is quite reasonable, given the orientation of the camera to the floor and since from the start of navigation, the system is assisting the visually impaired by avoiding obstacles.…”

Section: Reference Window Definitionmentioning

confidence: 99%

“…In Figure 3.a, a small central window in the bottom of the image is defined as a reference window with 25  25 pixels, which is assumed to be free from obstacles. This small area should represent the traversable area [14].…”

Section: Reference Window Definitionmentioning

confidence: 99%

“…Probability theory was the first model used to cope with value uncertainty problem, but they are not suitable to solve ambiguity categorizing problem [13]. Theories based on Fuzzy sets [14], proposed later on, by Zadeh, can model ambiguity but they fail in case of uncertain data. Theory of belief functions as well as possibility theory was proposed later on, to efficiently handle/model ambiguous/uncertain data [15].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Indoor/outdoor navigation system based on possibilistic traversable area segmentation for visually impaired people

Elleuch¹,

Sellami²,

Kallel³

2016

ELCVIA

View full text Add to dashboard Cite

Autonomous collision avoidance for visually impaired people requires a specific processing for an accurate definition of traversable area. Processing of a real time image sequence for traversable area segmentation is mandatory. Low cost systems suggest use of poor quality cameras. However, low cost camera suffers from great variability of traversable area appearance at indoor as well as outdoor environments. Taking into account ambiguity affecting object and traversable area appearance induced by reflections, illumination variations, occlusions (, etc...), an accurate segmentation of traversable area in such conditions remains a challenge. Moreover, at indoor and outdoor environments, more added variability is induced. In this paper, we present a fast traversable area segmentation approach from image sequence recorded by a low-cost monocular camera for navigation system. Taking into account all kinds of variability in the image, we apply possibility theory for modeling information ambiguity. An efficient way of updating the traversable area model in each environment condition is to consider traversable area samples from the same processed image for building its possibility maps. Then fusing these maps allows making a fair model definition of the traversable area. Performance of the proposed system was evaluated on public databases, at indoor and outdoor environments. Experimental results show that this method is challenging leading to higher segmentation rates.

show abstract

“…Capelle et al [43] coded every object pixel by means of complex tones which gradually change their tonal, timbric and binaural properties both in height and in azimuth, in order to allow the recognition of simple 2D graphic characters. Recently, [44] introduced a guidance system that detects people, faces and text, and [45] has presented a musical interface to provide obstacle information. In contrast to other references which also relay information to visually impaired people, CASBliP project is different because of the portability of the system, since all devices work onboard the blind person.…”

Section: Contextual Frameworkmentioning

confidence: 99%

Fuzzy Free Path Detection from Disparity Maps by Using Least-Squares Fitting to a Plane

Ortigosa

Morillas

2013

J Intell Robot Syst

View full text Add to dashboard Cite

A method to detect obstacle-free paths in real-time which works as part of a cognitive navigation aid system for visually impaired people is proposed. It is based on the analysis of disparity maps obtained from a stereo vision system which is carried by the blind user. The presented detection method consists of a fuzzy logic system that assigns a certainty to be part of a free path to each group of pixels, depending on the parameters of a planarmodel fitting. We also present experimental results on different real outdoor scenarios showing that our method is the most reliable in the sense that it minimizes the false positives rate.

show abstract

Image-to-MIDI mapping based on dynamic fuzzy color segmentation for visually impaired people

Cited by 9 publications

References 33 publications

Audio classification using braided convolutional neural networks

Audio classification using braided convolutional neural networks

Indoor/outdoor navigation system based on possibilistic traversable area segmentation for visually impaired people

Fuzzy Free Path Detection from Disparity Maps by Using Least-Squares Fitting to a Plane

Contact Info

Product

Resources

About