Zobrazeno 1 - 10
of 1 117
pro vyhledávání: '"Abbasnejad A"'
Autor:
Shamsi, Afshar, Becirovic, Rejisa, Argha, Ahmadreza, Abbasnejad, Ehsan, Alinejad-Rokny, Hamid, Mohammadi, Arash
Test time adaptation (TTA) equips deep learning models to handle unseen test data that deviates from the training distribution, even when source data is inaccessible. While traditional TTA methods often rely on entropy as a confidence metric, its eff
Externí odkaz:
http://arxiv.org/abs/2409.09251
Autor:
Cao, Haiyao, Zou, Jinan, Liu, Yuhang, Zhang, Zhen, Abbasnejad, Ehsan, Hengel, Anton van den, Shi, Javen Qinfeng
Accurately predicting stock returns is crucial for effective portfolio management. However, existing methods often overlook a fundamental issue in the market, namely, distribution shifts, making them less practical for predicting future markets or ne
Externí odkaz:
http://arxiv.org/abs/2409.00671
Autor:
Cao, Haiyao, Zhang, Zhen, Cai, Panpan, Liu, Yuhang, Zou, Jinan, Abbasnejad, Ehsan, Huang, Biwei, Gong, Mingming, Hengel, Anton van den, Shi, Javen Qinfeng
One of the significant challenges in reinforcement learning (RL) when dealing with noise is estimating latent states from observations. Causality provides rigorous theoretical support for ensuring that the underlying states can be uniquely recovered
Externí odkaz:
http://arxiv.org/abs/2408.13498
Autor:
Doan, Bao Gia, Nguyen, Dang Quang, Lindquist, Callum, Montague, Paul, Abraham, Tamas, De Vel, Olivier, Camtepe, Seyit, Kanhere, Salil S., Abbasnejad, Ehsan, Ranasinghe, Damith C.
Object detectors are vulnerable to backdoor attacks. In contrast to classifiers, detectors possess unique characteristics, architecturally and in task execution; often operating in challenging conditions, for instance, detecting traffic signs in auto
Externí odkaz:
http://arxiv.org/abs/2408.12122
Autor:
Doan, Bao Gia, Shamsi, Afshar, Guo, Xiao-Yu, Mohammadi, Arash, Alinejad-Rokny, Hamid, Sejdinovic, Dino, Ranasinghe, Damith C., Abbasnejad, Ehsan
Computational complexity of Bayesian learning is impeding its adoption in practical, large-scale tasks. Despite demonstrations of significant merits such as improved robustness and resilience to unseen or out-of-distribution inputs over their non- Ba
Externí odkaz:
http://arxiv.org/abs/2407.20891
Autor:
Zhang, Frederic Z., Albert, Paul, Rodriguez-Opazo, Cristian, Hengel, Anton van den, Abbasnejad, Ehsan
Pre-trained models produce strong generic representations that can be adapted via fine-tuning. The learned weight difference relative to the pre-trained model, known as a task vector, characterises the direction and stride of fine-tuning. The signifi
Externí odkaz:
http://arxiv.org/abs/2407.02880
Autor:
Rodriguez-Opazo, Cristian, Abbasnejad, Ehsan, Teney, Damien, Marrese-Taylor, Edison, Damirchi, Hamed, Hengel, Anton van den
Contrastive Language-Image Pretraining (CLIP) stands out as a prominent method for image representation learning. Various architectures, from vision transformers (ViTs) to convolutional networks (ResNets) have been trained with CLIP to serve as gener
Externí odkaz:
http://arxiv.org/abs/2405.17139
We study the unique, less-well understood problem of generating sparse adversarial samples simply by observing the score-based replies to model queries. Sparse attacks aim to discover a minimum number-the l0 bounded-perturbations to model inputs to c
Externí odkaz:
http://arxiv.org/abs/2404.05311
Autor:
Doan, Bao Gia, Nguyen, Dang Quang, Montague, Paul, Abraham, Tamas, De Vel, Olivier, Camtepe, Seyit, Kanhere, Salil S., Abbasnejad, Ehsan, Ranasinghe, Damith C.
The vulnerability of machine learning-based malware detectors to adversarial attacks has prompted the need for robust solutions. Adversarial training is an effective method but is computationally expensive to scale up to large datasets and comes at t
Externí odkaz:
http://arxiv.org/abs/2403.18309
Continual learning requires a model to adapt to ongoing changes in the data distribution, and often to the set of tasks to be performed. It is rare, however, that the data and task changes are completely unpredictable. Given a description of an overa
Externí odkaz:
http://arxiv.org/abs/2403.07356