Recent advances in unmanned aerial vehicle (UAV) technology combined with deep learning techniques have greatly improved agricultural monitoring. However, accurately processing images at low resolutions remains challenging for precision cultivation of succulents. To address this issue, this study proposes a novel method that combines cutting-edge super-resolution reconstruction (SRR) techniques with object detection and then applies the above model in a unified drone framework to achieve large-scale, reliable monitoring of succulent plants. Specifically, we introduce MambaIR, an innovative SRR method leveraging selective state-space models, significantly improving the quality of UAV-captured low-resolution imagery (achieving a PSNR of 23.83 dB and an SSIM of 79.60%) and surpassing current state-of-the-art approaches. Additionally, we develop Succulent-YOLO, a customized target detection model optimized for succulent image classification, achieving a mean average precision (mAP@50) of 87.8% on high-resolution images. The integrated use of MambaIR and Succulent-YOLO achieves an mAP@50 of 85.1% when tested on enhanced super-resolution images, closely approaching the performance on original high-resolution images. Through extensive experimentation supported by Grad-CAM visualization, our method effectively captures critical features of succulents, identifying the best trade-off between resolution enhancement and computational demands. By overcoming the limitations associated with low-resolution UAV imagery in agricultural monitoring, this solution provides an effective, scalable approach for evaluating succulent plant growth. Addressing image-quality issues further facilitates informed decision-making, reducing technical challenges. Ultimately, this study provides a robust foundation for expanding the practical use of UAVs and artificial intelligence in precision agriculture, promoting sustainable farming practices through advanced remote sensing technologies.

We have developed a system that streams and archives live sound from remote areas across Japan via an unmanned automatic camera. The system was used to carry out pilot bird censuses in woodland; this allowed us to examine the use of live sound transmission and the role of social media as a mediator in remote scientific monitoring. The system has been streaming sounds 8 h per day for more than five years. We demonstrated that: (1) the transmission of live sound from a remote woodland could be used effectively to monitor birds in a remote location; (2) the simultaneous involvement of several participants via Internet Relay Chat to listen to live sound transmissions could enhance the accuracy of census data collection; and (3) interactions through Twitter allowed members of the public to engage or help with the remote monitoring of birds and experience inaccessible nature through the use of novel technologies.

Urban time series data forecasting featuring significant contributions to sustainable development is widely studied as an essential task of the smart city. However, with the dramatic and rapid changes in the world environment, the assumption that data obey Independent Identically Distribution is undermined by the subsequent changes in data distribution, known as concept drift, leading to weak replicability and transferability of the model over unseen data. To address the issue, previous approaches typically retrain the model, forcing it to fit the most recent observed data. However, retraining is problematic in that it leads to model lag, consumption of resources, and model re-invalidation, causing the drift problem to be not well solved in realistic scenarios. In this study, we propose a new urban time series prediction model for the concept drift problem, which encodes the drift by considering the periodicity in the data and makes on-the-fly adjustments to the model based on the drift using a meta-dynamic network. Experiments on real-world datasets show that our design significantly outperforms state-of-the-art methods and can be well generalized to existing prediction backbones by reducing their sensitivity to distribution changes.

Ecocentric education programs should include a method for the in-depth understanding of multi-scale ecological time concepts. To accomplish this, the common restriction that ecocentric education should pertain only to realistic nature may have to be removed. The purpose of this research was to confirm the validity of a program featuring phenology observation, employing fixed-point time-lapse images as climate change learning, and to obtain suggestions on the influence of the program on the multi-timescale concepts of the learners. An observation sheet listing images of cherry flowering from 16 April to 15 March each year from 1996 to 2017 was created, and the 50-min educational program using the observation sheet was conducted with 189 third year junior high school students. The tendencies among students’ answers to the two questions before and after the program suggest that the program contributed to the students acquiring the hundreds-year timescale concept based on the short-term timescale concept of dynamic nature. The contribution of this research is to visualize long-term and multi-scale ecological time concepts. By combining long-term time-lapse images with everyday nature experiences, the possibility of expanding such emotions as wonder and attachment to nature towards a long-term ecological timescale is achieved.

In this study, we present a vision of how a human–computer–biosphere interaction (HCBI) can facilitate a sustainable society. HCBI extends and transforms the subject of human–computer interaction from countable people, objects, pets, and plants into an auditory biosphere that is an uncountable, a complex, and a non-linguistic soundscape. As an example, utilizing HCBI to experience forest soundscapes can help us feel one with nature, without physically being present in nature. The goal of HCBI is to achieve ecological interactions between humans and nature through computer systems without causing environmental destruction. To accomplish this, information connectivity must be created despite the physical separation between humans and the environment. This combination should also ensure ecological neutrality. In this paper, we present an overview of an HCBI concept, related work, methodologies, and developed interfaces. We used pre-recorded animal calls to enable a bio-acoustical feedback from the target wildlife. In this study, we primarily focus on the design and evaluation of a bio-acoustic interaction system utilizing tracking collars, microphones, speakers, infrared cameras, infrared heat sensors, micro-climate sensors, radio-tracking devices, GPS devices, radio clocks, embedded Linux boards, high-capacity batteries, and high-speed wireless communication devices. Our experiments successfully demonstrated bio-acoustic interactions between wildlife—more specifically, an endangered species of a wild cat—and human beings via a computer system, thus validating the HCBI concept.

Perceiving and modeling urban crowd movements are of great importance to smart city-related fields. Governments and public service operators can benefit from such efforts as they can be applied to crowd management, resource scheduling, and early emergency warning. However, most prior research on urban crowd modeling has failed to describe the dynamics and continuity of human mobility, leading to inconsistent and irrelevant results when they tackle multiple homogeneous forecasting tasks as they can only be modeled independently. To overcome this drawback, we propose to model human mobility from a new perspective, which uses the citywide crowd transition process constituted by a series of transition matrices from low order to high order, to help us understand how the crowd dynamics evolve step-by-step. We further propose a Deep Transition Process Network to process and predict such new high-dimensional data, where novel grid embedding with Graph Convolutional Network, parameter-shared Convolutional LSTM, and High-Dimensional Attention mechanism are designed to learn the complicated dependencies in terms of spatial, temporal, and ordinal features. We conduct experiments on two datasets generated by a large amount of GPS data collected from a real-world smartphone application. The experiment results demonstrate the superior performance of our proposed methodology over existing approaches.