The objective of the multi-condition human motion synthesis task is to
i...
The sample selection approach is very popular in learning with noisy lab...
This paper studies a new problem, active learning with partial labels
(A...
Generating unlabeled data has been recently shown to help address the
fe...
In recent years, research on learning with noisy labels has focused on
d...
In conventional supervised classification, true labels are required for
...
Although powerful graph neural networks (GNNs) have boosted numerous
rea...
Training a classifier exploiting a huge amount of supervised data is
exp...
Out-of-distribution (OOD) detection is an indispensable aspect of secure...
Label-noise learning (LNL) aims to increase the model's generalization g...
Adversarial detection aims to determine whether a given sample is an
adv...
Human motion generation aims to produce plausible human motion sequences...
Real-time emotion-based music arrangement, which aims to transform a giv...
Adversarial training (AT) is a robust learning algorithm that can defend...
Domain generalization (DG) aims to tackle the distribution shift between...
A common explanation for the failure of out-of-distribution (OOD)
genera...
Sketch-guided image editing aims to achieve local fine-tuning of the ima...
Zero-shot quantization (ZSQ) is promising for compressing and accelerati...
Robust generalization aims to tackle the most challenging data distribut...
Outlier exposure (OE) is powerful in out-of-distribution (OOD) detection...
Weakly Supervised Semantic Segmentation (WSSS) with image-level labels h...
Federated Semi-Supervised Learning (FSSL) aims to learn a global model f...
Privacy and security concerns in real-world applications have led to the...
Learning with noisy labels has become imperative in the Big Data era, wh...
By focusing on immersive interaction among users, the burgeoning Metaver...
Semantic-driven 3D shape generation aims to generate 3D objects conditio...
Dance-driven music generation aims to generate musical pieces conditione...
One-shot neural architecture search (NAS) substantially improves the sea...
Adversarial training (AT) with imperfect supervision is significant but
...
Supervised learning aims to train a classifier under the assumption that...
Adversarial training (AT) is proved to reliably improve network's robust...
Black-box attacks can generate adversarial examples without accessing th...
Robust learning on noisy-labeled data has been an important task in real...
Deep neural networks (DNNs) are found to be vulnerable to adversarial no...
Point cloud video transmission is challenging due to high encoding/decod...
Machine learning models are vulnerable to Out-Of-Distribution (OOD) exam...
On-device machine learning enables the lightweight deployment of
recomme...
We consider stochastic convex optimization for heavy-tailed data with th...
Robust overfitting widely exists in adversarial training of deep network...
Despite the success of invariant risk minimization (IRM) in tackling the...
The AutoAttack (AA) has been the most reliable method to evaluate advers...
Through using only a well-trained classifier, model-inversion (MI) attac...
In federated learning (FL), model performance typically suffers from cli...
In label-noise learning, estimating the transition matrix has attracted ...
Most recent self-supervised learning methods learn visual representation...
Due to the success of Graph Neural Networks (GNNs) in learning from
grap...
Overfitting widely exists in adversarial robust training of deep network...
Fair machine learning aims to avoid treating individuals or sub-populati...
Self-supervised learning has achieved a great success in the representat...
Federated learning (FL) aims at training a global model on the server si...