Zobrazeno 1 - 10
of 126
pro vyhledávání: '"Fahim, Faisal"'
Traffic forecasting in Intelligent Transportation Systems (ITS) is vital for intelligent traffic prediction. Yet, ITS often relies on data from traffic sensors or vehicle devices, where certain cities might not have all those smart devices or enablin
Externí odkaz:
http://arxiv.org/abs/2410.15589
Autor:
Ahmed, Sk Miraj, Niloy, Fahim Faisal, Chang, Xiangyu, Raychaudhuri, Dripta S., Oymak, Samet, Roy-Chowdhury, Amit K.
Adapting to dynamic data distributions is a practical yet challenging task. One effective strategy is to use a model ensemble, which leverages the diverse expertise of different models to transfer knowledge to evolving data distributions. However, th
Externí odkaz:
http://arxiv.org/abs/2401.02561
Source-Free Online Domain Adaptive Semantic Segmentation of Satellite Images under Image Degradation
Online adaptation to distribution shifts in satellite image segmentation stands as a crucial yet underexplored problem. In this paper, we address source-free and online domain adaptation, i.e., test-time adaptation (TTA), for satellite images, with t
Externí odkaz:
http://arxiv.org/abs/2401.02113
Autor:
Islam, Md Shazid, Nag, Sayak, Dutta, Arindam, Ahmed, Miraj, Niloy, Fahim Faisal, Roy-Chowdhury, Amit K.
Unsupervised domain adaptive segmentation typically relies on self-training using pseudo labels predicted by a pre-trained network on an unlabeled target dataset. However, the noisy nature of such pseudo-labels presents a major bottleneck in adapting
Externí odkaz:
http://arxiv.org/abs/2312.05407
Autor:
Niloy, Fahim Faisal, Ahmed, Sk Miraj, Raychaudhuri, Dripta S., Oymak, Samet, Roy-Chowdhury, Amit K.
Traditional test-time adaptation (TTA) methods face significant challenges in adapting to dynamic environments characterized by continuously changing long-term target distributions. These challenges primarily stem from two factors: catastrophic forge
Externí odkaz:
http://arxiv.org/abs/2311.04991
Existing high-resolution satellite image forgery localization methods rely on patch-based or downsampling-based training. Both of these training methods have major drawbacks, such as inaccurate boundaries between pristine and forged regions, the gene
Externí odkaz:
http://arxiv.org/abs/2307.11052
Reliable forecasting of traffic flow requires efficient modeling of traffic data. Indeed, different correlations and influences arise in a dynamic traffic network, making modeling a complicated task. Existing literature has proposed many different me
Externí odkaz:
http://arxiv.org/abs/2212.04548
Conventional forgery localizing methods usually rely on different forgery footprints such as JPEG artifacts, edge inconsistency, camera noise, etc., with cross-entropy loss to locate manipulated regions. However, these methods have the disadvantage o
Externí odkaz:
http://arxiv.org/abs/2210.02182
Autor:
Md Rezaul Hoque Khan, Md Sanowar Hosen, Atiqul Alam Chowdhury, Mohammad Rakibul Islam, Fahim Faisal, Mirza Muntasir Nishat, Nafiz Imtiaz Bin Hamid
Publikováno v:
Sensing and Bio-Sensing Research, Vol 45, Iss , Pp 100579- (2024)
The distillation method for extracting anhydrous ethanol from the volatile benzene water requires continuous supervision in the petro-chemical industry. This paper describes the design process and numerical analyses of a PCF sensor with a hollow core
Externí odkaz:
https://doaj.org/article/17afed7435434ac59abf34a86d7d6689
Decomposition of the evidence lower bound (ELBO) objective of VAE used for density estimation revealed the deficiency of VAE for representation learning and suggested ways to improve the model. In this paper, we investigate whether we can get similar
Externí odkaz:
http://arxiv.org/abs/2108.12734