Vinayak Abrol
YOU?
Author Swipe
View article: Evaluating Generative Models via Cubical Homology based Persistent Entropy
Evaluating Generative Models via Cubical Homology based Persistent Entropy Open
View article: Data encoding for healthcare data democratization and information leakage prevention
Data encoding for healthcare data democratization and information leakage prevention Open
View article: On Characterizing the Evolution of Embedding Space of Neural Networks using Algebraic Topology
On Characterizing the Evolution of Embedding Space of Neural Networks using Algebraic Topology Open
We study how the topology of feature embedding space changes as it passes through the layers of a well-trained deep neural network (DNN) through Betti numbers. Motivated by existing studies using simplicial complexes on shallow fully conne…
View article: Data Encoding For Healthcare Data Democratisation and Information Leakage Prevention
Data Encoding For Healthcare Data Democratisation and Information Leakage Prevention Open
The lack of data democratization and information leakage from trained models hinder the development and acceptance of robust deep learning-based healthcare solutions. This paper argues that irreversible data encoding can provide an effecti…
View article: Data Encoding For Healthcare Data Democratisation and Information Leakage Prevention
Data Encoding For Healthcare Data Democratisation and Information Leakage Prevention Open
The lack of data democratization and information leakage from trained models hinder the development and acceptance of robust deep learning-based healthcare solutions. This paper argues that irreversible data encoding can provide an effecti…
View article: Coordinate descent on the Stiefel manifold for deep neural network training
Coordinate descent on the Stiefel manifold for deep neural network training Open
To alleviate the cost incurred by orthogonality constraints in optimization and model training, we propose a stochastic coordinate descent algorithm on the Stiefel manifold.We compute expressions for geodesics on the Stiefel manifold with …
View article: Coordinate Descent on the Orthogonal Group for Recurrent Neural Network Training
Coordinate Descent on the Orthogonal Group for Recurrent Neural Network Training Open
We address the poor scalability of learning algorithms for orthogonal recurrent neural networks via the use of stochastic coordinate descent on the orthogonal group, leading to a cost per iteration that increases linearly with the number o…
View article: Activation function design for deep networks: linearity and effective initialisation
Activation function design for deep networks: linearity and effective initialisation Open
View article: Neural Transferability: Current Pitfalls and Striving for Optimal Scores
Neural Transferability: Current Pitfalls and Striving for Optimal Scores Open
View article: Coordinate descent on the orthogonal group for recurrent neural network training
Coordinate descent on the orthogonal group for recurrent neural network training Open
We propose to use stochastic Riemannian coordinate descent on the orthogonal group for recurrent neural network training. The algorithm rotates successively two columns of the recurrent matrix, an operation that can be efficiently implemen…
View article: An empirical study of derivative-free-optimization algorithms for targeted black-box attacks in deep neural networks
An empirical study of derivative-free-optimization algorithms for targeted black-box attacks in deep neural networks Open
View article: Activation function design for deep networks: linearity and effective\n initialisation
Activation function design for deep networks: linearity and effective\n initialisation Open
The activation function deployed in a deep neural network has great influence\non the performance of the network at initialisation, which in turn has\nimplications for training. In this paper we study how to avoid two problems at\ninitiali…
View article: An Empirical Study of Derivative-Free-Optimization Algorithms for\n Targeted Black-Box Attacks in Deep Neural Networks
An Empirical Study of Derivative-Free-Optimization Algorithms for\n Targeted Black-Box Attacks in Deep Neural Networks Open
We perform a comprehensive study on the performance of derivative free\noptimization (DFO) algorithms for the generation of targeted black-box\nadversarial attacks on Deep Neural Network (DNN) classifiers assuming the\nperturbation energy …
View article: Improving Generative Modelling in VAEs Using Multimodal Prior
Improving Generative Modelling in VAEs Using Multimodal Prior Open
In this paper we propose a conditional generative modelling (CGM) approach for unsupervised disentangled representation learning using variational autoencoder (VAE). CGM employs a multimodal/categorical conditional prior distribution in th…
View article: A Model-Based Derivative-Free Approach to Black-Box Adversarial Examples: BOBYQA
A Model-Based Derivative-Free Approach to Black-Box Adversarial Examples: BOBYQA Open
We demonstrate that model-based derivative free optimisation algorithms can generate adversarial targeted misclassification of deep networks using fewer network queries than non-model-based methods. Specifically, we consider the black-box …
View article: Understanding and Visualizing Raw Waveform-Based CNNs
Understanding and Visualizing Raw Waveform-Based CNNs Open
Modeling directly raw waveforms through neural networks for speech processing is gaining more and more attention. Despite its varied success, a question that remains is: what kind of information are such neural networks capturing or learni…
View article: Landuse impact on soil physical variability and erodibility in North Western subtropics of India
Landuse impact on soil physical variability and erodibility in North Western subtropics of India Open
The study was conducted to determine the impact of landuse on soil physical properties and erodibility.Representative soil samples were collected from surface and sub-surface soil depths.Soil physical properties and erodibility indices viz…
View article: Conv-codes: Audio Hashing For Bird Species Classification
Conv-codes: Audio Hashing For Bird Species Classification Open
In this work, we propose a supervised, convex representation based audio hashing framework for bird species classification. The proposed framework utilizes archetypal analysis, a matrix factorization technique, to obtain convex-sparse repr…
View article: Local compressed convex spectral embedding for bird species identification
Local compressed convex spectral embedding for bird species identification Open
This paper proposes a multi-layer alternating sparse−dense framework for bird species identification. The framework takes audio recordings of bird vocalizations and produces compressed convex spectral embeddings (CCSE). Temporal and freque…
View article: Archetypal Analysis Based Sparse Convex Sequence Kernel For Bird Activity Detection
Archetypal Analysis Based Sparse Convex Sequence Kernel For Bird Activity Detection Open
Publication in the conference proceedings of EUSIPCO, Kos island, Greece, 2017
View article: Class Specific Gmm Based Sparse Feature For Speech Units Classification
Class Specific Gmm Based Sparse Feature For Speech Units Classification Open
Publication in the conference proceedings of EUSIPCO, Kos island, Greece, 2017
View article: Gradient-based spectral visualization of CNNs using raw waveforms
Gradient-based spectral visualization of CNNs using raw waveforms Open
Modeling directly raw waveform through neural networks for speech processing is gaining more and more attention. Despite its varied success, a question that remains is: what kind of information are such neural networks capturing or learnin…
View article: Compressed sensing for unit selection based speech synthesis
Compressed sensing for unit selection based speech synthesis Open
Publication in the conference proceedings of EUSIPCO, Nice, France, 2015
View article: Making sense of randomness: an approach for fast recovery of compressively sensed signals
Making sense of randomness: an approach for fast recovery of compressively sensed signals Open
In compressed sensing (CS) framework, a signal is sampled below Nyquist rate, and the acquired compressed samples are generally random in nature. However, for efficient estimation of the actual signal, the sensing matrix must preserve the …