Anil Kumar Sao

Bio: Anil Kumar Sao is an academic researcher from Indian Institute of Technology Mandi. The author has contributed to research in topics: Sparse approximation & Face (geometry). The author has an hindex of 15, co-authored 79 publications receiving 696 citations. Previous affiliations of Anil Kumar Sao include Indian Institute of Technology Madras.

Solar radiation forecasting with multiple parameters neural networks

Abstract: Neural networks with a good modeling capability have been used increasingly to predict and forecast solar radiation. Even diverse application of neural network has been reported in literatures such as robotics, pattern recognition, forecasting, power systems, optimization and social/psychological sciences etc. The models have categorized the review under three major performance schemes such as delay, number of neurons and activation function for establishment of neural network architecture. In each of these categories, we summarize the major applications of eight well recognized and often used neural network models of which the last two are custom based. The anticipated model are initiated and validated with 10 metrological parameters further in sub-categories. Evaluation of its accuracy associated with special flexibility of the model is demonstrated through the results based on parameter range. In summary, we conclude the best result showing that the delays, neuron, transfer function, model, parameters and RMSE errors are in range of 15 or 30, 10 or 20, tansig, Elman Back Propagation network, bulb point temperature or direct normal radiation, 9–10 and 25–35% training to the test cases. The review discloses the incredible view of using the neural networks in solar forecast. The work of other researchers in the field of renewable energy and other energy systems is also reported which can be used in the future in the works of this field.

A syllable-based framework for unit selection synthesis in 13 Indian languages

Abstract: In this paper, we discuss a consortium effort on building text to speech (TTS) systems for 13 Indian languages. There are about 1652 Indian languages. A unified framework is therefore attempted required for building TTSes for Indian languages. As Indian languages are syllable-timed, a syllable-based framework is developed. As quality of speech synthesis is of paramount interest, unit-selection synthesizers are built. Building TTS systems for low-resource languages requires that the data be carefully collected an annotated as the database has to be built from the scratch. Various criteria have to addressed while building the database, namely, speaker selection, pronunciation variation, optimal text selection, handling of out of vocabulary words and so on. The various characteristics of the voice that affect speech synthesis quality are first analysed. Next the design of the corpus of each of the Indian languages is tabulated. The collected data is labeled at the syllable level using a semiautomatic labeling tool. Text to speech synthesizers are built for all the 13 languages, namely, Hindi, Tamil, Marathi, Bengali, Malayalam, Telugu, Kannada, Gujarati, Rajasthani, Assamese, Manipuri, Odia and Bodo using the same common framework. The TTS systems are evaluated using degradation Mean Opinion Score (DMOS) and Word Error Rate (WER). An average DMOS score of ≈3.0 and an average WER of about 20 % is observed across all the languages.

Deep-Sparse-Representation-Based Features for Speech Recognition

Abstract: Features derived using sparse representation (SR)-based approaches have been shown to yield promising results for speech recognition tasks. In most of the approaches, the SR corresponding to speech signal is estimated using a dictionary, which could be either exemplar based or learned. However, a single-level decomposition may not be suitable for the speech signal, as it contains complex hierarchical information about various hidden attributes. In this paper, we propose to use a multilevel decomposition (having multiple layers), also known as the deep sparse representation (DSR), to derive a feature representation for speech recognition. Instead of having a series of sparse layers, the proposed framework employs a dense layer between two sparse layers, which helps in efficient implementation. Our studies reveal that the representations obtained at different sparse layers of the proposed DSR model have complimentary information. Thus, the final feature representation is derived after concatenating the representations obtained at the sparse layers. This results in a more discriminative representation, and improves the speech recognition performance. Since the concatenation results in a high-dimensional feature, principal component analysis is used to reduce the dimension of the obtained feature. Experimental studies demonstrate that the proposed feature outperforms existing features for various speech recognition tasks.

CNN based segmentation of nuclei in PAP-smear images with selective pre-processing

Abstract: Cervical cancer is the second most common cause of death among women worldwide, but it can be treated if detected early. However, due to inter and intra observer variability in manual screening, automating the process is need of the hour. For classifying the cervical cells as normal vs abnormal, segmentation of nuclei as well as cytoplasm is a prerequisite. But the segmentation of nuclei is relatively more reliable and equally efficient for classification to that of cytoplasm. Hence, this paper proposes a new approach for segmentation of nuclei based on selective pre-processing and then passing the image patches to respective deep CNN (trained with/without pre-processed images) for pixel-wise 3 class labelling as nucleus, edge or background. We argue and demonstrate that a single pre-processing approach may not suit all images, as there are significant variations in nucleus sizes and chromatin patterns. The selective pre-processing is carried out to effectively address this issue. This also enables the deep CNNs to be better trained in spite of relatively less data, and thus better exploit the capability of CNN of good quality segmentation. The results show that the approach is effective for segmentation of nuclei in PAP-smears with an F-score of 0.90 on Herlev dataset as opposed to the without selective pre-processing F-scores of 0.78 (without pre-processing) and 0.82 (with pre-processing). The results also show the importance of considering 3 classes in CNN instead of 2 (nucleus and background) where the latter achieves an F-score as low as 0.63.

A Systematic Review of Compressive Sensing: Concepts, Implementations and Applications

Abstract: Compressive Sensing (CS) is a new sensing modality, which compresses the signal being acquired at the time of sensing. Signals can have sparse or compressible representation either in original domain or in some transform domain. Relying on the sparsity of the signals, CS allows us to sample the signal at a rate much below the Nyquist sampling rate. Also, the varied reconstruction algorithms of CS can faithfully reconstruct the original signal back from fewer compressive measurements. This fact has stimulated research interest toward the use of CS in several fields, such as magnetic resonance imaging, high-speed video acquisition, and ultrawideband communication. This paper reviews the basic theoretical concepts underlying CS. To bridge the gap between theory and practicality of CS, different CS acquisition strategies and reconstruction approaches are elaborated systematically in this paper. The major application areas where CS is currently being used are reviewed here. This paper also highlights some of the challenges and research directions in this field.

Review of distributed generation (DG) system planning and optimisation techniques: Comparison of numerical and mathematical modelling methods

Abstract: An overview of numerical and mathematical modelling-based distributed generation (DG) system optimisation techniques is presented in this review paper. The objective is to compare different aspects of these two broad classes of DG optimisation techniques, explore their applications, and identify potential research directions from reviewed studies. Introductory descriptions of general electrical power system and DG system are first provided, followed by reviews on renewable resource assessment, load demand analysis, model formulation, and optimisation techniques. In renewable resource assessment model review, uncertain solar and wind energy resources are emphasised whereas applications of forecasting models have been highlighted based on their prediction horizons, computational power requirement, and training data intensity. For DG optimisation framework, (solar, wind and tidal) power generator, energy storage and energy balance models are discussed; in optimisation technique section, both numerical and mathematical modelling optimisation methods are reviewed, analysed and criticised with recommendations for their improvements. In overall, this review provides preliminary guidelines, research gaps and recommendations for developing a better and more user-friendly DG energy planning optimisation tool.