College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Sliding mode control & Control theory. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

Noise Estimation and Noise Removal Techniques for Speech Recognition in Adverse Environment

Abstract: Noise is ubiquitous in almost all acoustic environments. The speech signal, that is recorded by a microphone is generally infected by noise originating from various sources. Such contamination can change the characteristics of the speech signals and degrade the speech quality and intelligibility, thereby causing significant harm to human-to-machine communication systems.

A Method for Melanoma Skin Cancer Detection Using Dermoscopy Images

Abstract: Now days, Skin cancer is life threatening disease which causes human death. Abnormal growth of melanocytic cells causes a skin cancer. Due to malignancy feature skin cancer is also known as melanoma. Melanoma appears on the skin due to exposure of ultraviolet radiation and genetic factors. So melanoma lesion appears as black or brown in colour. Early detection of melanoma can cure completely. Biopsy is a traditional method for detecting skin cancer. This method is painful and invasive. This method requires laboratory testing so it is time consuming. Therefore, in order to solve the above stated issues computer aided diagnosis for skin cancer is needed. Computer aided diagnosis uses Dermoscopy for capturing the skin image. In this paper first pre-processing of the skin image is done. After pre-processing lesion part is segmented by using image segmentation technique which is followed by feature extraction in which unique features are extracted from segmented lesion. After feature extraction, classification by using support vector machine is performed for classifying the skin image as normal skin and melanoma skin cancer. The proposed system results shows that support vector machine with linear kernel gives optimum accuracy.

Automated region-based hybrid compression for digital imaging and communications in medicine magnetic resonance imaging images for telemedicine applications

Abstract: Many classes of images contains spatial regions which are more important than other regions. Compression methods which are capable of delivering higher reconstruction quality are attractive in this situation for the important parts. For the medical images, only a small portion of the image might be diagnostically useful, but the cost of a wrong interpretation is high. Hence, Region Based Coding (RBC) technique is significant for medical image compression and transmission. Lossless compression in these `regions` and lossy compression for rest of image can helps to achieve high efficiency and performance in telemedicine applications. This paper proposes an automated, efficient and low complexity, lossless, scalable RBC for Digital Imaging and Communications in Medicine (DICOM) images. The advantages of RBC are exploited in this paper, segmenting the region into various regions of importance and subjecting varying bit-rates for optimal performance. Moreover, the combined effects of Integer Wavelet Transform (IWT) and bit-rate limiting compression technique for lesser important regions helps reconstruct the image, reversibly, up to a desired quality. The overall compression thus reaches a satisfactory level to be able to safely transmit the image in limited bandwidth over a telemedicine network and reconstruct diagnostic details for treatment, most faithfully.

Devanagari Handwritten Character Recognition using fine-tuned Deep Convolutional Neural Network on trivial dataset

Abstract: In order to rapidly build an automatic and precise system for image recognition and categorization, deep learning is a vital technology. Handwritten character classification also gaining more attention due to its major contribution in automation and specially to develop applications for helping visually impaired people. Here, the proposed work highlighting on fine-tuning approach and analysis of state-of-the-art Deep Convolutional Neural Network (DCNN) designed for Devanagari Handwritten characters classification. A new Devanagari handwritten characters dataset is generated which is publicly available. Datasets consist of total 5800 isolated images of 58 unique character classes: 12 vowels, 36 consonants and 10 numerals. In addition to this database, a two-stage VGG16 deep learning model is implemented to recognize those characters using two advanced adaptive gradient methods. A two-stage approach of deep learning is developed to enhance overall success of the proposed Devanagari Handwritten Character Recognition System (DHCRS). The first model achieves 94.84% testing accuracy with training loss of 0.18 on new dataset. Moreover, the second fine-tuned model requires very fewer trainable parameters and notably less training time to achieve state-of-the-art performance on a very small dataset. It achieves 96.55% testing accuracy with training loss of 0.12. We also tested the proposed model on four different benchmark datasets of isolated characters as well as digits of Indic scripts. For all the datasets tested, we achieved the promising results.