Jesse Berent
6 papers · 2019–2023 · 5 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+3 more ↓ Show less ↑
π£ Hot Topic Early Bird π Cross-Pollinator (10) πΊοΈ Taxonomy Completionist (15) π Interdisciplinary Bridge π§ Keyword Pioneer
π
Conference Polyglot
(5)
π
Triple Crown
β
The Questioner
Conferences
ICML (2)
CVPR (1)
ICCV (1)
ICLR (1)
NIPS (1)
Top co-authors
Keywords
image classification
(3)
label noise
(3)
privileged information
(2)
object detection
(1)
weakly supervised learning
(1)
covariance matrix
(1)
weight sharing
(1)
latent variable
(1)
vision-language model
(1)
clean sample selection
(1)
pretrained model
(1)
training datum
(1)
aleatoric uncertainty
(1)
heteroscedastic noise
(1)
noisy datum
(1)
image-level annotation
(1)
neural network
(1)
clean datum
(1)
caption supervision
(1)
representation learning
(1)
Papers
Three Towers: Flexible Contrastive Learning with Pretrained Image Models
NIPS 2023
Massively Scaling Heteroscedastic Classifiers
ICLR 2023
When does Privileged information Explain Away Label Noise?
ICML 2023
Transfer and Marginalize: Explaining Away Label Noise with Privileged Information
ICML 2022
Correlated Input-Dependent Label Noise in Large-Scale Image Classification
CVPR 2021
Cap2Det: Learning to Amplify Weak Caption Supervision for Object Detection
ICCV 2019