Search for: All records

There is an urgent need to develop global observation networks to quantify biodiversity trends for evaluating achievements of targets of Kunming-Montreal Global Biodiversity Framework. Camera traps are a commonly used tool, with the potential to enhance global observation networks for monitoring wildlife population trends and has the capacity to constitute global observation networks by applying a unified sampling protocol. The Snapshot protocol is simple and easy for camera trapping which is applied in North America and Europe. However, there is no regional camera-trap network with the Snapshot protocol in Asia. We present the first dataset from a collaborative camera-trap survey using the Snapshot protocol in Japan conducted in 2023. We collected data at 90 locations across nine arrays for a total of 6162 trap-nights of survey effort. The total number of sequences with mammals and birds was 7967, including 20 mammal species and 23 avian species. Apart from humans, wild boar, sika deer and rodents were the most commonly observed taxa on the camera traps, covering 57.9% of all the animal individuals. We provide the dataset with a standard format of Wildlife Insights, but also with Camtrap DP 1.0 format. Our dataset can be used for a part of the global dataset for comparing relative abundances of wildlife and for a baseline of wildlife population trends in Japan. It can also used for training machine-learning models for automatic species identifications. 

Human mobility data offers valuable insights for many applications such as urban planning and pandemic response, but its use also raises privacy concerns. In this paper, we introduce the Hierarchical and Multi-Resolution Network (HRNet), a novel deep generative model specifically designed to synthesize realistic human mobility data while guaranteeing differential privacy. We first identify the key difficulties inherent in learning human mobility data under differential privacy. In response to these challenges, HRNet integrates three components: a hierarchical location encoding mechanism, multi-task learning across multiple resolutions, and private pre-training. These elements collectively enhance the model's ability under the constraints of differential privacy. Through extensive comparative experiments utilizing a real-world dataset, HRNet demonstrates a marked improvement over existing methods in balancing the utility-privacy trade-off. 

Differentially Private Federated Learning (DP-FL) has garnered attention as a collaborative machine learning approach that ensures formal privacy. Most DP-FL approaches ensure DP at the record-level within each silo for cross-silo FL. However, a single user's data may extend across multiple silos, and the desired user-level DP guarantee for such a setting remains unknown. In this study, we present Uldp-FL, a novel FL framework designed to guarantee user-level DP in cross-silo FL where a single user's data may belong to multiple silos. Our proposed algorithm directly ensures user-level DP through per-user weighted clipping, departing from group-privacy approaches. We provide a theoretical analysis of the algorithm's privacy and utility. Additionally, we improve the utility of the proposed algorithm with an enhanced weighting strategy based on user record distribution and design a novel private protocol that ensures no additional information is revealed to the silos and the server. Experiments on real-world datasets show substantial improvements in our methods in privacy-utility trade-offs under user-level DP compared to baseline methods. To the best of our knowledge, our work is the first FL framework that effectively provides user-level DP in the general cross-silo FL setting.