Publication

Training DNNs with Hybrid Block Floating Point

Publications associées (31)

VWR2A: A Very-Wide-Register Reconfigurable-Array Architecture for Low-Power Embedded Devices

David Atienza Alonso, Benoît Walter Denkinger, Miguel Peon Quiros

Edge-computing requires high-performance energy-efficient embedded systems. Fixed-function or custom accelerators, such as FFT or FIR filter engines, are very efficient at implementing a particular functionality for a given set of constraints. However, the ...

ASSOC COMPUTING MACHINERY2022

Beyond Time-Average Convergence: Near-Optimal Uncoupled Online Learning via Clairvoyant Multiplicative Weights Update

Efstratios Panteleimon Skoulakis

In this paper we provide a novel and simple algorithm, Clairvoyant Multiplicative Weights Updates (CMWU), for convergence to \textit{Coarse Correlated Equilibria} (CCE) in general games. CMWU effectively corresponds to the standard MWU algorithm but where ...

2022

A Shadow Perspective on Equivariant Hochschild Homologies

Kathryn Hess Bellwald, Inbar Klang

Shadows for bicategories, defined by Ponto, provide a useful framework that generalizes classical and topological Hochschild homology. In this paper, we define Hochschild-type invariants for monoids in a symmetric monoidal, simplicial model category V, as ...

OXFORD UNIV PRESS2022

Visual Focus of Attention Estimation in 3D Scene with an Arbitrary Number of Targets

Jean-Marc Odobez, Rémy Alain Siegfried

Visual Focus of Attention (VFOA) estimation in conversation is challenging as it relies on difficult to estimate information (gaze) combined with scene features like target positions and other contextual information (speaking status) allowing to disambigua ...

IEEE COMPUTER SOC2021

Masksembles for Uncertainty Estimation

Pascal Fua, Timur Bagautdinov, Pierre Bruno Baqué, Nikita Durasov

Deep neural networks have amply demonstrated their prowess but estimating the reliability of their predictions remains challenging. Deep Ensembles are widely considered as being one of the best methods for generating uncertainty estimates but are very expe ...

2021

Global bifurcation at isolated singular points of the Hadamard derivative

Charles Stuart

Consider F is an element of C(RxX,Y) such that F(lambda, 0) = 0 for all lambda is an element of R, where X and Y are Banach spaces. Bifurcation from the line Rx{0} of trivial solutions is investigated in cases where F(lambda, center dot ) need not be Frech ...

2021

Fourier Interpolation From Spheres

Martin Peter Stoller

In every dimension d >= 2, we give an explicit formula that expresses the values of any Schwartz function on R-d only in terms of its restrictions, and the restrictions of its Fourier transform, to all origin-centered spheres whose radius is the square roo ...

2021

Invariant Integrals on Topological Groups and Applications

Vasco Schiavo

We study a fixed point property for linear actions of discrete groups on weakly complete convex proper cones in locally convex topological vector spaces. We search to understand the class of discrete groups which enjoys this property and we try to generali ...

EPFL2021

Sparsely Observed Functional Time Series: Theory and Applications

Tomas Rubin

Functional time series is a temporally ordered sequence of not necessarily independent random curves. While the statistical analysis of such data has been traditionally carried out under the assumption of completely observed functional data, it may well ha ...

EPFL2021

Equinox: Training (for Free) on a Custom Inference Accelerator

Babak Falsafi, Ahmet Caner Yüzügüler, Mario Paulo Drumond Lages De Oliveira, Arash Pourhabibi Zarandi, Louis Coulon, Martin Jaggi

DNN inference accelerators executing online services exhibit low average loads because of service demand variability, leading to poor resource utilization. Unfortunately, reclaiming idle inference cycles is difficult as other workloads can not execute on a ...

ACM2021

Semantic Perturbations with Normalizing Flows for Improved Generalization

Tatjana Chavdarova, Martin Jaggi, Sebastian Urban Stich

Data augmentation is a widely adopted technique for avoiding overfitting when training deep neural networks. However, this approach requires domain-specific knowledge and is often limited to a fixed set of hard-coded transformations. Recently, several work ...

IEEE2021