Searcharxiv⌕ Search

arXiv subjects

Yujun Huang

Publications and source records attributed to Yujun Huang.

21 records · Page 2Linked to original sources

Adapter Learning in Pretrained Feature Extractor for Continual Learning of Diseases

Currently intelligent diagnosis systems lack the ability of continually learning to diagnose new diseases once deployed, under the condition of preserving old disease knowledge. In particular, updating an intelligent diagnosis system with training data of new diseases would cause catastrophic forgetting of old disease knowledge. To address the catastrophic forgetting issue, an Adapter-based Continual Learning framework called ACL is proposed to help effectively learn a set of new diseases at each round (or task) of continual learning, without changing the shared feature extractor. The learnable lightweight task-specific adapter(s) can be flexibly designed (e.g., two convolutional layers) and then added to the pretrained and fixed feature extractor. Together with a specially designed task-specific head which absorbs all previously learned old diseases as a single "out-of-distribution" category, task-specific adapter(s) can help the pretrained feature extractor more effectively extract discriminative features between diseases. In addition, a simple yet effective fine-tuning is applied to collaboratively fine-tune multiple task-specific heads such that outputs from different heads are comparable and consequently the appropriate classifier head can be more accurately selected during model inference. Extensive empirical evaluations on three image datasets demonstrate the superior performance of ACL in continual learning of new diseases. The source code is available at https://github.com/GiantJun/CL_Pytorch.

cs.CV↗

Increasing and Diverging Greenhouse Gas Emissions of Urban Wastewater Treatment in China

Upgrading effluent standards of wastewater treatment plants (WWTPs) and repairing sewerage systems leads to contradictions and synergies between water pollution control and climate change mitigation. This affects historical trajectories and characteristics of greenhouse gas (GHG) emissions from China's WWTPs, which stay inadequately studied. Here we establish emissions inventories of China's WWTPs using plant-level WWTP operational data. We find that removed amount of chemical oxygen demand and ammonia nitrogen increased 0.8 and 1.3 times during 2009-2019, while WWTP GHG emissions increased 1.8 times, being 6 times national GHG emissions growth rate. Increasing sludge yield and electricity intensity became primary driving factors in 2015 because of stricter effluent standards and lower influent contaminant concentration. We defined Functional Unit-Gini coefficient to quantify divergence of WWTP GHG emissions, which grew from 0.20 in 2009 to 0.29 in 2019. Diversified sludge disposal methods and energy structure increased the inequality, while upgrading effluent standards decreased it.

physics.soc-ph↗

Learned Distributed Image Compression with Multi-Scale Patch Matching in Feature Domain

Beyond achieving higher compression efficiency over classical image compression codecs, deep image compression is expected to be improved with additional side information, e.g., another image from a different perspective of the same scene. To better utilize the side information under the distributed compression scenario, the existing method (Ayzik and Avidan 2020) only implements patch matching at the image domain to solve the parallax problem caused by the difference in viewing points. However, the patch matching at the image domain is not robust to the variance of scale, shape, and illumination caused by the different viewing angles, and can not make full use of the rich texture information of the side information image. To resolve this issue, we propose Multi-Scale Feature Domain Patch Matching (MSFDPM) to fully utilizes side information at the decoder of the distributed image compression model. Specifically, MSFDPM consists of a side information feature extractor, a multi-scale feature domain patch matching module, and a multi-scale feature fusion network. Furthermore, we reuse inter-patch correlation from the shallow layer to accelerate the patch matching of the deep layer. Finally, we nd that our patch matching in a multi-scale feature domain further improves compression rate by about 20% compared with the patch matching method at image domain (Ayzik and Avidan 2020).

cs.CV↗