Search Results for author:
Found 14 papers, 9 papers with code
LEA: Improving Sentence Similarity Robustness to Typos Using Lexical Attention Bias
We show that this is also the case for sentence similarity, a fundamental task in multiple domains, e. g. matching, retrieval or paraphrasing.
Block-SCL: Blocking Matters for Supervised Contrastive Learning in Product Matching
Product matching is a fundamental step for the global understanding of consumer behavior in e-commerce.
Domain Randomization for Object Counting
We introduce a domain randomization approach for object counting based on synthetic datasets that are quick and inexpensive to generate.
How Important is Importance Sampling for Deep Budgeted Training?
We suggest that, given a specific budget, the best course of action is to disregard the importance and introduce adequate data augmentation; e. g. when reducing the budget to a 30% in CIFAR-10/100, RICAP data augmentation maintains accuracy, while importance sampling does not.
Addressing out-of-distribution label noise in webly-labelled data
We propose a simple solution to bridge the gap with a fully clean dataset using Dynamic Softening of Out-of-distribution Samples (DSOS), which we design on corrupted versions of the CIFAR-100 dataset, and compare against state-of-the-art algorithms on the web noise perturbated MiniImageNet and Stanford datasets and on real label noise datasets: WebVision 1. 0 and Clothing1M.
The Importance of Importance Sampling for Deep Budgeted Training
For example, training in CIFAR-10/100 with 30% of the full training budget, a uniform sampling strategy with certain data augmentation surpasses the performance of 100% budget models trained with standard data augmentation.
Multi-Objective Interpolation Training for Robustness to Label Noise
We further propose a novel label noise detection method that exploits the robust feature representations learned via contrastive learning to estimate per-sample soft-labels whose disagreements with the original labels accurately identify noisy samples.
Ranked #21 on
Image Classification
on mini WebVision 1.0
Unsupervised Contrastive Learning of Sound Event Representations
Self-supervised representation learning can mitigate the limitations in recognition tasks with few manually labeled data but abundant unlabeled data---a common scenario in sound event research.
Reliable Label Bootstrapping for Semi-Supervised Learning
We propose Reliable Label Bootstrapping (ReLaB), an unsupervised preprossessing algorithm which improves the performance of semi-supervised algorithms in extremely low supervision settings.
Toward the Automatic Retrieval and Annotation of Outsider Art images: A Preliminary Statement
The aim of this position paper is to establish an initial approach to the automatic classification of digital images about the Outsider Art style of painting.
Towards Robust Learning with Different Label Noise Distributions
However, we show that different noise distributions make the application of this trick less straightforward and propose to continuously relabel all images to reveal a discriminative loss against multiple distributions.
Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learning
In the context of image classification, recent advances to learn from unlabeled samples are mainly focused on consistency regularization methods that encourage invariant predictions for different perturbations of unlabeled samples.
Unsupervised Label Noise Modeling and Loss Correction
Specifically, we propose a beta mixture to estimate this probability and correct the loss by relying on the network prediction (the so-called bootstrapping loss).
Ranked #44 on
Image Classification
on Clothing1M
On guiding video object segmentation
This guiding process relies on foreground masks from independent algorithms (i. e. state-of-the-art algorithms) to implement an attention mechanism that incorporates the spatial location of foreground and background to compute their separated representations.
