MD
Matthijs Douze
cs.CVcs.LGcs.CLcs.AIcs.CRcs.DBcs.IRcs.DSstat.MLcs.MM
On Valency
published · living versionsW_dby3ngaf·v1 · currentpublished
FastText.zip: Compressing text classification models
with Armand Joulin, Edouard Grave, Piotr Bojanowski, Hérve Jégou +1
1 version
Preprints & journals
55 papers in the corpus · 2011–2025Short window attention enables long-term memorization2509.24552v3 · Loic Cabannes, Maximilian Beck, Gergely Szilvasy et al.2025 · 0 citationsarXiv
The Faiss library2401.08281v4 · Matthijs Douze, Alexandr Guzhva, Chengqi Deng et al.2024 · 76 citationsarXiv
Watermark Anything with Localized Messages2411.07231v2 · Tom Sander, Pierre Fernandez, Alain Durmus et al.2024 · 1 citationarXiv
Inference-time sparse attention with asymmetric indexing2502.08246v2 · Pierre-Emmanuel Mazar'e, Gergely Szilvasy, Maria Lomeli et al.2025 · 0 citationsarXiv
Machine learning and high dimensional vector search2502.16931v1 · Matthijs Douze2025 · 0 citationsarXiv
Lossless Compression of Vector IDs for Approximate Nearest Neighbor Search2501.10479v1 · Daniel Severo, Giuseppe Ottaviano, Matthew Muckley et al.2025 · 0 citationsarXiv
Qinco2: Vector Compression and Search with Improved Implicit Neural Codebooks2501.03078v1 · Th'eophane Vallaeys, Matthew Muckley, Jakob Verbeek et al.2025 · 0 citationsarXiv
Watermarking Makes Language Models Radioactive2402.14904v2 · Tom Sander, Pierre Fernandez, Alain Durmus et al.2024 · 7 citationsarXiv
Results of the Big ANN: NeurIPS'23 competition2409.17424v1 · Harsha Vardhan Simhadri, Martin Aumuller, Amir Ingber et al.2024 · 1 citationarXiv
Residual Quantization with Implicit Neural Codebooks2401.14732v2 · Iris A. M. Huijben, Matthijs Douze, Matthew Muckley et al.2024 · 0 citationsarXiv
Vector search with small radiuses2403.10746v1 · Gergely Szilvasy, Pierre-Emmanuel Mazar'e, Matthijs Douze2024 · 0 citationsarXiv
Functional Invariants to Watermark Large Transformers2310.11446v2 · Pierre Fernandez, Guillaume Couairon, Teddy Furon et al.2023 · 7 citationsarXiv
DeDrift: Robust Similarity Search under Content Drift2308.02752v1 · Dmitry Baranchuk, Matthijs Douze, Yash Upadhyay et al.2023 · 10 citationsarXiv
The Stable Signature: Rooting Watermarks in Latent Diffusion Models2303.15435v2 · Pierre Fernandez, Guillaume Couairon, Herv'e J'egou et al.2023 · 219 citationsarXiv
The 2023 Video Similarity Dataset and Challenge2306.09489v1 · Ed Pizzi, Giorgos Kordopatis-Zilos, Hiral Patel et al.2023 · 16 citationsarXiv
Generalized Few-Shot Video Classification With Video Retrieval and Feature Generation.34652997 · Xian, Yongqin, Korbar, Bruno, Douze, Matthijs et al.2022 · 22 citationsIEEE transactions on pattern analysis and machine intelligence. 2022;44(12):8949-8961
Embedding Arithmetic of Multimodal Queries for Image Retrieval2112.03162v2 · Guillaume Couairon, Matthieu Cord, Matthijs Douze et al.2021 · 19 citationsarXiv
Active Image Indexing2210.10620v1 · Pierre Fernandez, Matthijs Douze, Herv'e J'egou et al.2022 · 0 citationsarXiv
Results of the NeurIPS'21 Challenge on Billion-Scale Approximate Nearest Neighbor Search2205.03763v1 · Harsha Vardhan Simhadri, George Williams, Martin Aumuller et al.2022 · 14 citationsarXiv
A Self-Supervised Descriptor for Image Copy Detection2202.10261v2 · Ed Pizzi, Sreya Dutta Roy, Sugosh Nagavara Ravindra et al.2022 · 90 citationsarXiv
Watermarking Images in Self-Supervised Latent Spaces2112.09581v2 · Pierre Fernandez, Alexandre Sablayrolles, Teddy Furon et al.2021 · 97 citationsarXiv
Nearest neighbor search with compact codes: A decoder perspective2112.09568v2 · Kenza Amara, Matthijs Douze, Alexandre Sablayrolles et al.2021 · 2 citationsarXiv
The 2021 Image Similarity Dataset and Challenge2106.09672v4 · Matthijs Douze, Giorgos Tolias, Ed Pizzi et al.2021 · 4 citationsarXiv
Results and findings of the 2021 Image Similarity Challenge2202.04007v1 · Zoe Papakipos, Giorgos Tolias, Tomas Jenicek et al.2022 · 1 citationarXiv
Fixing the train-test resolution discrepancy1906.06423v4 · Hugo Touvron, Andrea Vedaldi, Matthijs Douze et al.2019 · 23 citationsarXiv
Generalized Few-Shot Video Classification with Video Retrieval and Feature Generation2007.04755v2 · Yongqin Xian, Bruno Korbar, Matthijs Douze et al.2020 · 22 citationsarXiv
XCiT: Cross-Covariance Image Transformers2106.09681v2 · Alaaeldin El-Nouby, Hugo Touvron, Mathilde Caron et al.2021 · 234 citationsarXiv
LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference2104.01136v2 · Ben Graham, Alaaeldin El-Nouby, Hugo Touvron et al.2021 · 929 citationsarXiv
Training data-efficient image transformers & distillation through attention2012.12877v2 · Hugo Touvron, Matthieu Cord, Matthijs Douze et al.2020 · 1,219 citationsarXiv
Grafit: Learning fine-grained image representations with coarse labels2011.12982v1 · Hugo Touvron, Alexandre Sablayrolles, Matthijs Douze et al.2020 · 77 citationsarXiv
Fixing the train-test resolution discrepancy: FixEfficientNet2003.08237v5 · Hugo Touvron, Andrea Vedaldi, Matthijs Douze et al.2020 · 65 citationsarXiv
Powers of layers for image-to-image translation2008.05763v1 · Hugo Touvron, Matthijs Douze, Matthieu Cord et al.2020 · 5 citationsarXiv
Data Augmenting Contrastive Learning of Speech Representations in the Time Domain2007.00991v1 · Eugene Kharitonov, Morgane Riviere, Gabriel Synnaeve et al.2020 · 114 citationsarXiv
Radioactive data: tracing through training2002.00937v1 · Alexandre Sablayrolles, Matthijs Douze, Cordelia Schmid et al.2020 · 35 citationsarXiv
Lead2Gold: Towards exploiting the full potential of noisy transcriptions for speech recognition1910.07323v1 · Adrien Dufraux, Emmanuel Vincent, Awni Hannun et al.2019 · 9 citationsarXiv
Spreading vectors for similarity search1806.03198v3 · Alexandre Sablayrolles, Matthijs Douze, Cordelia Schmid et al.2018 · 56 citationsarXiv
White-box vs Black-box: Bayes Optimal Strategies for Membership Inference1908.11229v1 · Alexandre Sablayrolles, Matthijs Douze, Yann Ollivier et al.2019 · 111 citationsarXiv
MultiGrain: a unified image embedding for classes and instances1902.05509v2 · Maxim Berman, Herv'e J'egou, Andrea Vedaldi et al.2019 · 33 citationsarXiv
Deep Clustering for Unsupervised Learning of Visual Features1807.05520v2 · Mathilde Caron, Piotr Bojanowski, Armand Joulin et al.2018 · 2,840 citationsarXiv
D\'ej\`a Vu: an empirical evaluation of the memorization properties of ConvNets1809.06396v1 · Alexandre Sablayrolles, Matthijs Douze, Cordelia Schmid et al.2018 · 9 citationsarXiv
Low-shot learning with large-scale diffusion1706.02332v3 · Matthijs Douze, Arthur Szlam, Bharath Hariharan et al.2017 · 133 citationsarXiv
Polysemous codes1609.01882v2 · Matthijs Douze, Herv'e J'egou, Florent Perronnin2016 · 0 citationsarXiv
Billion-scale similarity search with GPUs1702.08734v1 · Jeff Johnson, Matthijs Douze, Herv'e J'egou2017 · 3,867 citationsarXiv
Link and code: Fast indexing with graphs and compact regression codes1804.09996v2 · Matthijs Douze, Alexandre Sablayrolles, Herv'e J'egou2018 · 41 citationsarXiv
How should we evaluate supervised hashing?1609.06753v3 · Alexandre Sablayrolles, Matthijs Douze, Herv'e J'egou et al.2016 · 101 citationsarXiv
An evaluation of large-scale methods for image instance and class discovery1708.02898v1 · Matthijs Douze, Herv'e J'egou, Jeff Johnson2017 · 7 citationsarXiv
Learning Joint Multilingual Sentence Representations with Neural Machine Translation1704.04154v2 · Holger Schwenk, Matthijs Douze2017 · 207 citationsarXiv
QuickCSG: Fast Arbitrary Boolean Combinations of N Solids1706.01558v1 · Matthijs Douze, Jean-S'ebastien Franco, Bruno Raffin2017 · 2 citationsarXiv
FastText.zip: Compressing text classification models1612.03651v1 · Armand Joulin, Edouard Grave, Piotr Bojanowski et al.2016 · 893 citationsarXivon Valency
Convolutional Patch Representations for Image Retrieval: an Unsupervised Approach1603.00438v1 · Mattis Paulin (LEAR), Julien Mairal (LEAR), Matthijs Douze (LEAR) et al.2016 · 66 citationsarXiv
Career total: 146 works. 55 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.