2 December 2021
Theoretical and experimental findings to improve regression applications: a 3D rotation case study. Code.
2021
29 November 2021
A novel, plug and play model for human 3D shape estimation of the body or hands, in videos which is trained by mimicking the BERT algorithm from the natural language processing community.
2021
4 November 2021
How to improve the inference speed of large Multilingual NTM models and do fast and parameter-efficient domain and language adaptation with them. Work published at EMNLP and WMT 2021.
2021
29 September 2021
A novel approach to indoor localization that uses magnetic field data from smartphone sensors and deep learning
2021
30 August 2021
A new approach to learning few-shot imitation agents whereby you simply feed demonstrations of a new test task to the learned policy called DCRL. This new approach has several advantages.
2021
3 August 2021
Podcast & edited transcript on Energy Based Models (EBMs). Guests, Hady Elsahar and Marc Dymetman work on EBMs in the field of natural language and co-organised the ICLR 2021 workshop on EBMs.
2021
15 July 2021
A new approach to learning few-shot imitation agents whereby you simply feed demonstrations of a new test task to the learned policy called DCRL. This new approach has several advantages.
2021
8 July 2021
A new sparse bi-encoder BERT-based model for effective and efficient first-stage ranking. The first to rival dense models.
2021
21 June 2021
Using domain randomization and meta-learning, computer vision models forget less when exposed to training samples from new domains. Remembering is a crucial element in the deployment of self-driving cars and robots which interact in dynamic environments.
2021
19 May 2021
NAVER LABS releases world's biggest visual localization dataset of indoor spaces with over 130K images. Dataset built with NAVER LABS mapping robots M1X & COMET and available in unified data format kapture.
2021
14 May 2021
Using a novel algorithm we explore how effectively a single policy, learned by reinforcement learning, can modulate robot behaviour from risk-averse to risk-neutral, so that robots can safely navigate everyday environments like homes and shops.
2021
30 April 2021
A novel framework for controlled NLG called 'Generation with Distributional Control', achieves great generality on the types of constraints that can be imposed and has a large potential to remedy the problem of bias in language models.
2021