Publications associées à Single-Handed Eyes-Free Universal Input Device

VETIM: Expanding the Vocabulary of Text-to-Image Models only with Text

Sabine Süsstrunk, Radhakrishna Achanta, Mahmut Sami Arpa, Martin Nicolas Everaert

Text-to-image models, such as Stable Diffusion, can generate high-quality images from simple textual prompts. With methods such as Textual Inversion, it is possible to expand the vocabulary of these models with additional concepts, by learning the vocabula ...

BMVA2023

Hidden Markov Modeling Over Graphs

Ali H. Sayed, Stefan Vlaski, Virginia Bordignon, Mert Kayaalp

This work proposes a multi-agent filtering algorithm over graphs for finite-state hidden Markov models (HMMs), which can be used for sequential state estimation or for tracking opinion formation over dynamic social networks. We show that the difference fro ...

IEEE2022

A note on universal inference

Anthony Christopher Davison, Timmy Rong Tian Tse

Universal inference enables the construction of confidence intervals and tests without regularity conditions by splitting the data into two parts and appealing to Markov's inequality. Previous investigations have shown that the cost of this generality is a ...

WILEY2022

Product of experts for robot learning from demonstration

Emmanuel Pignat

Adaptability and ease of programming are key features necessary for a wider spread of robotics in factories and everyday assistance. Learning from demonstration (LfD) is an approach to address this problem. It aims to develop algorithms and interfaces such ...

EPFL2021

On quantifying the quality of acoustic models in hybrid DNN-HMM ASR

Hervé Bourlard, Afsaneh Asaei, Pranay Dighe

We propose an information theoretic framework for quantitative assessment of acoustic models used in hidden Markov model (HMM) based automatic speech recognition (ASR). The HMM backend expects that (i) the acoustic model yields accurate state conditional e ...

ELSEVIER2020

Multilingual Training and Adaptation in Speech Recognition

Sibo Tong

State-of-the-art acoustic models for Automatic Speech Recognition (ASR) are based on Hidden Markov Models (HMM) and Deep Neural Networks (DNN) and often require thousands of hours of transcribed speech data during training. Therefore, building multilingual ...

EPFL2020

Gaussian channel transmission of images and audio files using cryptcoding

Vladimir Ilievski

Random codes based on quasigroups (RCBQ) are cryptcodes, i.e. they are error-correcting codes, which provide information security. Cut-Decoding and 4-Sets-Cut-Decoding algorithms for these codes are defined elsewhere. Also, the performance of these codes f ...

2019

Low-rank tensor methods for large Markov chains and forward feature selection methods

Francisco Santos Paredes Quartin de Macedo

In the first part of this thesis, we present and compare several approaches for the determination of the steady-state of large-scale Markov chains with an underlying low-rank tensor structure. Such structure is, in our context of interest, associated with ...

EPFL2018

Neural encoding of auditory features during music perception and imagery

José del Rocio Millán Ruiz, Stéphanie Martin

Despite many behavioral and neuroimaging investigations, it remains unclear how the human cortex represents spectrotemporal sound features during auditory imagery, and how this representation compares to auditory perception. To assess this, we recorded ele ...

2018

Efficient Learning from Comparisons

Lucas Maystre

Humans are comparison machines: comparing and choosing an item among a set of alternatives (such as objects or concepts) is arguably one of the most natural ways for us to express our preferences and opinions. In many applications, the analysis of data con ...

EPFL2018

Multigrid Methods Combined With Low-Rank Approximation For Tensor-Structured Markov Chains

Daniel Kressner, Francisco Santos Paredes Quartin de Macedo

Markov chains that describe interacting subsystems suffer from state space explosion but lead to highly structured matrices. In this work, we propose a novel tensor-based algorithm to address such tensor-structured Markov chains. Our algorithm combines a t ...

2018

Low-rank and Sparse Soft Targets to Learn Better DNN Acoustic Models

Hervé Bourlard, Afsaneh Asaei, Pranay Dighe

Conventional deep neural networks (DNN) for speech acoustic modeling rely on Gaussian mixture models (GMM) and hidden Markov model (HMM) to obtain binary class labels as the targets for DNN training. Subword classes in speech recognition systems correspond ...

Ieee2017

Learning Stable Task Sequences from Demonstration with Linear Parameter Varying Systems and Hidden Markov Models

Aude Billard, José Ramon Medina Hernandez

The problem of acquiring multiple tasks from demonstration is typi- cally divided in two sequential processes: (1) the segmentation or identification of different subgoals/subtasks and (2) a separate learning process that parameterizes a control policy for ...

robot-learning2017

High-Speed Wireline Link Design

Hazar Yüksel

High-speed serial links are a crucial application of semiconductor technology and have been the enabler of the scaling of computing systems. The increasing data-rate requirements of these links have only been partially satisfied by advancements in process ...

EPFL2017

Exploiting Low-dimensional Structures to Enhance DNN based Acoustic Modeling in Speech Recognition

Hervé Bourlard, Afsaneh Asaei, Pranay Dighe

We propose to model the acoustic space of deep neural network (DNN) class-conditional posterior probabilities as a union of low- dimensional subspaces. To that end, the training posteriors are used for dictionary learning and sparse coding. Sparse represen ...

IEEE2016

A Cache-Assisted Scratchpad Memory for Multiple-Bit-Error Correction

Nooshin Sadat Mirzadeh, Hossein Asadi

Scratchpad memory (SPM) is widely used in modern embedded processors to overcome the limitations of cache memory. The high vulnerability of SPM to soft errors, however, limits its usage in safety-critical applications. This paper proposes an efficient faul ...

Ieee-Inst Electrical Electronics Engineers Inc2016

Energy vs. Reliability Trade-offs Exploration in Biomedical Ultra-Low Power Devices

David Atienza Alonso, Pablo Garcia del Valle, Andreas Peter Burg, Loris Gérard Duch, Shrikanth Ganapathy

State-of-the-art wearable devices such as embedded biomedical monitoring systems apply voltage scaling to lower as much as possible their energy consumption and achieve longer battery lifetimes. While embedded memories often rely on Error Correction Codes ...

Ieee2016

Finding Steady States of Communicating Markov Processes Combining Aggregation/Disaggregation with Tensor Techniques

Francisco Santos Paredes Quartin de Macedo

Stochastic models for interacting processes feature a dimensionality that grows exponentially with the number of processes. This state space explosion severely impairs the use of standard methods for the numerical analysis of such Markov chains. In this wo ...

Springer Int Publishing Ag2016

HMM-Based Error Correction Mechanism for Five-Key Chording Keyboards

Bixio Rimoldi, Pierre Dillenbourg, Adrian Dan Tarniceriu

As different text input devices lead to different typing error patterns, considering the device characteristics when designing an error correction mechanism can lead to significantly improved results. In this paper, we propose and evaluate a spelling corre ...

2015

Mitigating the Impact of Faults in Unreliable Memories For Error-Resilient Applications

Andreas Peter Burg, Georgios Karakonstantis, Adam Shmuel Teman, Shrikanth Ganapathy

Inherently error-resilient applications in areas such as signal processing, machine learning and data analytics provide opportunities for relaxing reliability requirements, and thereby reducing the overheads incurred by conventional error correction scheme ...

2015