Christian Raymond
YOU?
Author Swipe
View article: Meta-Learning Loss Functions for Deep Neural Networks
Meta-Learning Loss Functions for Deep Neural Networks Open
Humans can often quickly and efficiently solve new complex learning tasks given only a small set of examples. In contrast, modern artificially intelligent systems often require thousands or millions of observations in order to solv…
View article: Meta-Learning Loss Functions for Deep Neural Networks
Meta-Learning Loss Functions for Deep Neural Networks Open
Humans can often quickly and efficiently solve complex new learning tasks given only a small set of examples. In contrast, modern artificially intelligent systems often require thousands or millions of observations in order to solve even t…
View article: Meta-Learning Neural Procedural Biases
Meta-Learning Neural Procedural Biases Open
The goal of few-shot learning is to generalize and achieve high performance on new unseen learning tasks, where each task has only a limited number of examples available. Gradient-based meta-learning attempts to address this challenging ta…
View article: Fast and Efficient Local Search for Genetic Programming Based Loss Function Learning
Fast and Efficient Local Search for Genetic Programming Based Loss Function Learning Open
In this paper, we develop upon the topic of loss function learning, an emergent meta-learning paradigm that aims to learn loss functions that significantly improve the performance of the models trained under them. Specifically, we propose …
View article: StyleGAN-based heatmap generator for face alignment with limited training data
StyleGAN-based heatmap generator for face alignment with limited training data Open
While the performance of face alignment models has been improving over the years, they still need large, annotated datasets during their training to perform well. In this paper, we propose a new architecture to perform face alignment with …
View article: Meta-Learning Adaptive Loss Functions
Meta-Learning Adaptive Loss Functions Open
Loss function learning is a new meta-learning paradigm that aims to automate the essential task of designing a loss function for a machine learning model. Existing techniques for loss function learning have shown promising results, often i…
View article: Qui de DrBERT, Wikipédia ou Flan-T5 s'y connaît le plus en questions médicales ?
Qui de DrBERT, Wikipédia ou Flan-T5 s'y connaît le plus en questions médicales ? Open
International audience
View article: Learning Symbolic Model-Agnostic Loss Functions via Meta-Learning
Learning Symbolic Model-Agnostic Loss Functions via Meta-Learning Open
In this paper, we develop upon the emerging topic of loss function learning, which aims to learn loss functions that significantly improve the performance of the models trained under them. Specifically, we propose a new meta-learning frame…
View article: Exploring StyleGAN Latent Space for Face Alignment with Limited Training Data
Exploring StyleGAN Latent Space for Face Alignment with Limited Training Data Open
View article: Can We Predict How Challenging Spoken Language Understanding Corpora Are Across Sources, Languages, and Domains?
Can We Predict How Challenging Spoken Language Understanding Corpora Are Across Sources, Languages, and Domains? Open
View article: SCAF: Skip-Connections in Auto-encoder for Face Alignment with Few Annotated Data
SCAF: Skip-Connections in Auto-encoder for Face Alignment with Few Annotated Data Open
View article: Dialogue History Integration into End-to-End Signal-to-Concept Spoken Language Understanding Systems
Dialogue History Integration into End-to-End Signal-to-Concept Spoken Language Understanding Systems Open
This work investigates the embeddings for representing dialog history in\nspoken language understanding (SLU) systems. We focus on the scenario when the\nsemantic information is extracted directly from the speech signal by means of a\nsing…
View article: The classification of schizophrenia based on brain structural features: A machine learning approach
The classification of schizophrenia based on brain structural features: A machine learning approach Open
将机器学习应用于精神疾患的临床和基础研究是近年来的趋势。研究者将机器学习应用于精神分裂症患者及高危人群的T1加权像和弥散张量成像的脑影像数据中,为了解疾病的生理病理学机制提供帮助。回顾以往研究发现额叶及颞叶的脑结构特征具有较高的区分能力,行为数据和脑影像数据结合的分类效果优于单模态数据。现阶段研究存在样本量不足和泛化能力欠缺的局限,未来研究应注意扩大样本量、制定标准化的分类方法,从而进一步探究机器学习在精神疾患中的作用。
View article: Mining Polysemous Triplets with Recurrent Neural Networks for Spoken Language Understanding
Mining Polysemous Triplets with Recurrent Neural Networks for Spoken Language Understanding Open
View article: Benchmarking benchmarks: introducing new automatic indicators for benchmarking Spoken Language Understanding corpora
Benchmarking benchmarks: introducing new automatic indicators for benchmarking Spoken Language Understanding corpora Open
International audience
View article: Bidirectional deep architecture for Arabic speech recognition
Bidirectional deep architecture for Arabic speech recognition Open
Nowadays, the real life constraints necessitates controlling modern machines using human intervention by means of sensorial organs. The voice is one of the human senses that can control/monitor modern interfaces. In this context, Automatic…
View article: IRISA at SMM4H 2018: Neural Network and Bagging for Tweet Classification
IRISA at SMM4H 2018: Neural Network and Bagging for Tweet Classification Open
International audience
View article: Heat Map Based Feature Ranker: In depth comparison with popular methods
Heat Map Based Feature Ranker: In depth comparison with popular methods Open
International audience
View article: Is ATIS too shallow to go deeper for benchmarking Spoken Language Understanding models?
Is ATIS too shallow to go deeper for benchmarking Spoken Language Understanding models? Open
International audience
View article: Participation de l’IRISA à DeFT 2018 : classification et annotation d’opinion dans des tweets
Participation de l’IRISA à DeFT 2018 : classification et annotation d’opinion dans des tweets Open
National audience
View article: Bi-directional recurrent end-to-end neural network classifier for spoken Arab digit recognition
Bi-directional recurrent end-to-end neural network classifier for spoken Arab digit recognition Open
International audience
View article: A Crossmodal Approach to Multimodal Fusion in Video Hyperlinking
A Crossmodal Approach to Multimodal Fusion in Video Hyperlinking Open
International audience
View article: IRISA at SMM4H 2018: Neural Network and Bagging for Tweet Classification
IRISA at SMM4H 2018: Neural Network and Bagging for Tweet Classification Open
This paper describes the systems developed by IRISA to participate to the four tasks of the SMM4H 2018 challenge. For these tweet classification tasks, we adopt a common approach based on recurrent neural networks (BiLSTM). Our main contri…
View article: Label-dependency coding in Simple Recurrent Networks for Spoken Language Understanding
Label-dependency coding in Simple Recurrent Networks for Spoken Language Understanding Open
International audience
View article: Participation de l'IRISA à DeFT2017 : systèmes de classification de complexité croissante
Participation de l'IRISA à DeFT2017 : systèmes de classification de complexité croissante Open
National audience
View article: Generative Adversarial Networks for Multimodal Representation Learning in Video Hyperlinking
Generative Adversarial Networks for Multimodal Representation Learning in Video Hyperlinking Open
International audience
View article: Generative Adversarial Networks for Multimodal Representation Learning in Video Hyperlinking
Generative Adversarial Networks for Multimodal Representation Learning in Video Hyperlinking Open
Continuous multimodal representations suitable for multimodal information retrieval are usually obtained with methods that heavily rely on multimodal autoencoders. In video hyperlinking, a task that aims at retrieving video segments, the s…
View article: One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network
One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network Open
There is an inherent need for autonomous cars, drones, and other robots to have a notion of how their environment behaves and to anticipate changes in the near future. In this work, we focus on anticipating future appearance given the curr…
View article: One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network
One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network Open
View article: Exploiting Multimodality in Video Hyperlinking to Improve Target Diversity
Exploiting Multimodality in Video Hyperlinking to Improve Target Diversity Open