Translating Embeddings for Modeling Multi-relational Data
- Antoine Bordes, Nicolas Usunier, Alberto García-Durán, J. Weston, Oksana Yakhnenko
- Computer ScienceNIPS
- 5 December 2013
TransE is proposed, a method which models relationships by interpreting them as translations operating on the low-dimensional embeddings of the entities, which proves to be powerful since extensive experiments show that TransE significantly outperforms state-of-the-art methods in link prediction on two knowledge bases.
End-to-End Object Detection with Transformers
- Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, Sergey Zagoruyko
- Computer ScienceEuropean Conference on Computer Vision
- 26 May 2020
This work presents a new method that views object detection as a direct set prediction problem, and demonstrates accuracy and run-time performance on par with the well-established and highly-optimized Faster RCNN baseline on the challenging COCO object detection dataset.
Large-scale Simple Question Answering with Memory Networks
- Antoine Bordes, Nicolas Usunier, S. Chopra, J. Weston
- Computer ScienceArXiv
- 5 June 2015
This paper studies the impact of multitask and transfer learning for simple question answering; a setting for which the reasoning required to answer is quite easy, as long as one can retrieve the correct evidence given a question, which can be difficult in large-scale conditions.
WSABIE: Scaling Up to Large Vocabulary Image Annotation
- J. Weston, Samy Bengio, Nicolas Usunier
- Computer ScienceInternational Joint Conference on Artificial…
- 16 July 2011
This work proposes a strongly performing method that scales to image annotation datasets by simultaneously learning to optimize precision at the top of the ranked list of annotations for a given image and learning a low-dimensional joint embedding space for both images and annotations.
Parseval Networks: Improving Robustness to Adversarial Examples
- Moustapha Cissé, Piotr Bojanowski, Edouard Grave, Y. Dauphin, Nicolas Usunier
- Computer ScienceInternational Conference on Machine Learning
- 28 April 2017
It is shown that Parseval networks match the state-of-the-art in terms of accuracy on CIFAR-10/100 and Street View House Numbers while being more robust than their vanilla counterpart against adversarial examples.
Fader Networks: Manipulating Images by Sliding Attributes
- Guillaume Lample, Neil Zeghidour, Nicolas Usunier, Antoine Bordes, Ludovic Denoyer, M. Ranzato
- Computer ScienceNIPS
- 1 June 2017
A new encoder-decoder architecture that is trained to reconstruct images by disentangling the salient information of the image and the values of attributes directly in the latent space is introduced, which results in much simpler training schemes and nicely scales to multiple attributes.
Canonical Tensor Decomposition for Knowledge Base Completion
- Timothée Lacroix, Nicolas Usunier, G. Obozinski
- Computer ScienceInternational Conference on Machine Learning
- 19 June 2018
This work motivates and test a novel regularizer, based on tensor nuclear $p$-norms, and presents a reformulation of the problem that makes it invariant to arbitrary choices in the inclusion of predicates or their reciprocals in the dataset.
Improving Neural Language Models with a Continuous Cache
- Edouard Grave, Armand Joulin, Nicolas Usunier
- Computer ScienceInternational Conference on Learning…
- 4 November 2016
A simplified version of memory augmented networks, which stores past hidden activations as memory and accesses them through a dot product with the current hidden activation, which is very efficient and scales to very large memory sizes.
Music Source Separation in the Waveform Domain
- Alexandre D'efossez, Nicolas Usunier, L. Bottou, F. Bach
- Computer ScienceArXiv
- 25 September 2019
Demucs is proposed, a new waveform-to-waveform model, which has an architecture closer to models for audio generation with more capacity on the decoder, and human evaluations show that Demucs has significantly higher quality than Conv-Tasnet, but slightly more contamination from other sources, which explains the difference in SDR.
Tensor Decompositions for temporal knowledge base completion
- Timothée Lacroix, G. Obozinski, Nicolas Usunier
- Computer ScienceInternational Conference on Learning…
- 10 April 2020
A solution inspired by the canonical decomposition of tensors of order 4 is proposed for link prediction under temporal constraints and a new dataset for knowledge base completion constructed from Wikidata is proposed, larger than previous benchmarks by an order of magnitude, as a new reference for evaluating temporal and non-temporal link prediction methods.
...
...