Search

We target the problem named unsupervised domain adaptive semantic segmentation. A key in this campaign consists in reducing the domain shift, so that a classifier based on labeled data from one domain can generalize well to other domains. With the advancement of adversarial learning method, recent works prefer the strategy of aligning the marginal distribution in the feature spaces for minimizing the domain discrepancy. However, based on the observance in experiments, only focusing on aligning global marginal distribution but ignoring the local joint distribution alignment fails to be the optimal choice. Other than that, the noisy factors existing in the feature spaces, which are not relevant to the target task, entangle with the domain invariant factors improperly and make the domain distribution alignment more difficult. To address those problems, we introduce two new modules, Significance-aware Information Bottleneck (SIB) and Category-level alignment (CLA), to construct a purified embedding-based category-level adversarial network. As the name suggests, our designed network, CLAN, can not only disentangle the noisy factors and suppress their influences for target tasks but also utilize those purified features to conduct a more delicate level domain calibration, i.e., global marginal distribution and local joint distribution alignment simultaneously. In three domain adaptation tasks, i.e., GTA5 → Cityscapes, SYNTHIA → Cityscapes and Cross Season, we validate that our proposed method matches the state of the art in segmentation accuracy.

Around the 2010s, China's economy has entered a "new normal" stage-transitioning from an extensive to an intensive growth mode. This study aims to investigate whether China's energy and carbon footprints also show these "new normal" characteristics. We evaluate China's energy and carbon footprints of 42 sectors from 2007 to 2017. The "new normal" characteristics are reflected from three dimensions: trend, structure, and driving factor. The results show that while the growth rate of China's energy and carbon footprints has slowed down, the total footprints are still increasing. The footprints induced by consumption have gradually exceeded those induced by export, and the tertiary industrial sectors became critical nodes in footprint networks. Furthermore, economic structure and development level have been major drivers of energy and carbon footprint growth. The findings reveal that China's energy and carbon footprints show similar "new normal" characteristics as economic development. This supports the targeted formulation of China's future policies to achieve sustainable development., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2021 Elsevier B.V. All rights reserved.)

Chronology is a necessary component of paleoclimatology. Radiocarbon dating plays a central role in determining the ages of geological samples younger than ca. 50 ka BP. However, there are many limitations for its application, including radiocarbon reservoir effects, which may cause incorrect chronology in many lakes. Here we demonstrate temporal changes in the radiocarbon reservoir age of Lake Xingyun, Southwestern China, where radiocarbon ages based on bulk organic matter have been reported in previous studies. Our new radiocarbon ages, determined from terrestrial plant macrofossils suggest that the radiocarbon reservoir age changed from 960 to 2200 years during the last 8500 cal a BP years. These changes to the reservoir effect were associated with inputs from either pre-aged organic carbon or 14C-depleted hard water in Lake Xingyun caused by hydrological change in the lake system. The radiocarbon reservoir age may in return be a good indicator for the carbon source in lake ecosystems and depositional environment.

Geochemical measurements on well-dated sediment cores from Lake Er (Erhai) are used to determine the timing of changes in metal concentrations over 4500 years in Yunnan, a borderland region in southwestern China noted for rich mineral deposits but with inadequately documented metallurgical history. Our findings add new insight into the impacts and environmental legacy of human exploitation of metal resources in Yunnan history. We observe an increase in copper at 1500 BC resulting from atmospheric emissions associated with metallurgy. These data clarify the chronological issues related to links between the onset of Yunnan metallurgy and the advent of bronze technology in adjacent Southeast Asia, subjects that have been debated for nearly half a century. We also observe an increase from 1100 to 1300 AD in a number of heavy metals including lead, silver, zinc, and cadmium from atmospheric emissions associated with silver smelting. Culminating during the rule of the Mongols, known as the Yuan Dynasty (1271-1368 AD), these metal concentrations approach levels three to four times higher than those from industrialized mining activity occurring within the catchment today. Notably, the concentrations of lead approach levels at which harmful effects may be observed in aquatic organisms. The persistence of this lead pollution over time created an environmental legacy that likely contributes to known issues in modern day sediment quality. We demonstrate that historic metallurgical production in Yunnan can cause substantial impacts on the sediment quality of lake systems, similar to other paleolimnological findings around the globe.

Mobile Visual Location Recognition (MVLR) has attracted a lot of researchers' attention in the past few years. Existing MVLR applications commonly use Query-by-Example (QBE) based image retrieval principle to fulfill the location recognition task. However, the QBE framework is not reliable enough due to the variations in the capture conditions and viewpoint changes between the query image and the database images. To solve the above problem, we make following contributions to the design of a panorama based on-device MVLR system. Firstly, we design a heading (from digital compass) aware BOF (Bag-of-features) model to generate the descriptors of panoramic images. Our approach fully considers the characteristics of the panoramic images and can facilitate the panorama based on-device MVLR to a large degree. Secondly, to search high dimensional visual descriptors directly on mobile devices, we propose an effective bilinear compressed sensing based encoding method. While being fast and accurate enough for on-device implementation, our algorithm can also reduce the memory usage of projection matrix significantly. Thirdly, we also release a panoramas database as well as a set of test panoramic quires which can be used as a new benchmark to facilitate further research in the area. Experimental results prove the effectiveness of the proposed methods for on-device MVLR applications.

This paper focuses on online scene learning and fast camera relocalisation which are two key problems currently limiting the performance of wide area augmented reality systems. Firstly, we propose to use adaptive random trees to deal with the online scene learning problem. The algorithm can provide more accurate recognition rates than traditional methods, especially with large scale workspaces. Secondly, we use the enhanced PROSAC algorithm to obtain a fast camera relocalisation method. Compared with traditional algorithms, our method can significantly reduce the computation complexity, which facilitates to a large degree the process of online camera relocalisation. Finally, we implement our algorithms in a multithreaded manner by using a parallel-computing scheme. Camera tracking, scene mapping, scene learning and relocalisation are separated into four threads by using multi-CPU hardware architecture. While providing real-time tracking performance, the resulting system also possesses the ability to track multiple maps simultaneously. Some experiments have been conducted to demonstrate the validity of our methods.