CenterNet-Saccade: Enhancing Sonar Object Detection with Lightweight Global Feature Extraction

Sonar imaging technology is widely used in the field of marine and underwater monitoring because sound waves can be transmitted in elastic media, such as the atmosphere and seawater, without much interference. In underwater object detection, due to the unique characteristics of the monitored sonar image, and since the target in an image is often accompanied by its own shadow, we can use the relative relationship between the shadow and the target for detection. To make use of shadow-information-aided detection and realize accurate real-time detection in sonar images, we put forward a network based on a lightweight module. By using the attention mechanism with a global receptive field, the network can make the target pay attention to the shadow information in the global environment, and because of its exquisite design, the computational time of the network is greatly reduced. Specifically, we design a ShuffleBlock model adapted to Hourglass to make the backbone network lighter. The concept of CNN dimension reduction is applied to MHSA to make it more efficient while paying attention to global features. Finally, CenterNet's unreasonable distribution method of positive and negative samples is improved. Simulation experiments were carried out using the proposed sonar object detection dataset. The experimental results further verify that our improved model has obvious advantages over many existing conventional deep learning models. Moreover, the real-time monitoring performance of our proposed model is more conducive to the implementation in the field of ocean monitoring.

Toward a decade of ocean science for sustainable development through acoustic animal tracking

The ocean is a key component of the Earth's dynamics, providing a great variety of ecosystem services to humans. Yet, human activities are globally changing its structure and major components, including marine biodiversity. In this context, the United Nations has proclaimed a Decade of Ocean Science for Sustainable Development to tackle the scientific challenges necessary for a sustainable use of the ocean by means of the Sustainable Development Goal 14 (SDG14). Here, we review how Acoustic animal Tracking, a widely distributed methodology of tracking marine biodiversity with electronic devices, can provide a roadmap for implementing the major Actions to achieve the SDG14. We show that acoustic tracking can be used to reduce and monitor the effects of marine pollution including noise, light, and plastic pollution. Acoustic tracking can be effectively used to monitor the responses of marine biodiversity to human-made infrastructures and habitat restoration, as well as to determine the effects of hypoxia, ocean warming, and acidification. Acoustic tracking has been historically used to inform fisheries management, the design of marine protected areas, and the detection of essential habitats, rendering this technique particularly attractive to achieve the sustainable fishing and spatial protection target goals of the SDG14. Finally, acoustic tracking can contribute to end illegal, unreported, and unregulated fishing by providing tools to monitor marine biodiversity against poachers and promote the development of Small Islands Developing States and developing countries. To fully benefit from acoustic tracking supporting the SDG14 Targets, trans-boundary collaborative efforts through tracking networks are required to promote ocean information sharing and ocean literacy. We therefore propose acoustic tracking and tracking networks as relevant contributors to tackle the scientific challenges that are necessary for a sustainable use of the ocean promoted by the United Nations.

Field Testing of Biohybrid Robotic Jellyfish to Demonstrate Enhanced Swimming Speeds

Biohybrid robotic designs incorporating live animals and self-contained microelectronic systems can leverage the animals’ own metabolism to reduce power constraints and act as natural chassis and actuators with damage tolerance. Previous work established that biohybrid robotic jellyfish can exhibit enhanced speeds up to 2.8 times their baseline behavior in laboratory environments. However, it remains unknown if the results could be applied in natural, dynamic ocean environments and what factors can contribute to large animal variability. Deploying this system in the coastal waters of Massachusetts, we validate and extend prior laboratory work by demonstrating increases in jellyfish swimming speeds up to 2.3 times greater than their baseline, with absolute swimming speeds up to 6.6 ± 0.3 cm s⁻¹. These experimental swimming speeds are predicted using a hydrodynamic model with morphological and time-dependent input parameters obtained from field experiment videos. The theoretical model can provide a basis to choose specific jellyfish with desirable traits to maximize enhancements from robotic manipulation. With future work to increase maneuverability and incorporate sensors, biohybrid robotic jellyfish can potentially be used to track environmental changes in applications for ocean monitoring.

Autonomous and In Situ Ocean Environmental Monitoring on Optofluidic Platform

Determining the distributions and variations of chemical elements in oceans has significant meanings for understanding the biogeochemical cycles, evaluating seawater pollution, and forecasting the occurrence of marine disasters. The primary chemical parameters of ocean monitoring include nutrients, pH, dissolved oxygen (DO), and heavy metals. At present, ocean monitoring mainly relies on laboratory analysis, which is hindered in applications due to its large size, high power consumption, and low representative and time-sensitive detection results. By integrating photonics and microfluidics into one chip, optofluidics brings new opportunities to develop portable microsystems for ocean monitoring. Optofluidic platforms have advantages in respect of size, cost, timeliness, and parallel processing of samples compared with traditional instruments. This review describes the applications of optofluidic platforms on autonomous and in situ ocean environmental monitoring, with an emphasis on their principles, sensing properties, advantages, and disadvantages. Predictably, autonomous and in situ systems based on optofluidic platforms will have important applications in ocean environmental monitoring.

Partnering with Fishing Fleets to Monitor Ocean Conditions

Engaging ocean users, including fishing fleets, in oceanographic and ecological research is a valuable method for collecting high-quality data, improving cost efficiency, and increasing societal appreciation for scientific research. As research partners, fishing fleets provide broad access to and knowledge of the ocean, and fishers are highly motivated to use the data collected to better understand the ecosystems in which they harvest. Here, we discuss recent trends in collaborative research that have increased the capacity of and access to scientific data collection. We also describe common elements of successful collaborative research programs, including definition of a scientific problem and goals, choice of technology, data collection and sampling design, data management and dissemination, and data analysis and communication. Finally, we review four case studies that demonstrate the general principles of effective collaborative research as well as the utility of citizen-collected data for academic research and fisheries management. We also discuss the challenge of funding, particularly as it relates to maintaining collaborative research programs in the long term. We conclude with a discussion of likely future trends. Ultimately, we predict that collaborative research will continue to grow in importance as climate change increasingly impacts ocean ecosystems, commercial fisheries, and the global food supply.

Detection of climate change-driven trends in phytoplankton phenology

The timing of the annual phytoplankton spring bloom is likely to be altered in response to climate change. Quantifying that response has, however, been limited by the typically coarse temporal resolution (monthly) of global climate models. Here, we use higher resolution model output (maximum 5 days) to investigate how phytoplankton bloom timing changes in response to projected 21st century climate change, and how the temporal resolution of data influences the detection of long-term trends. We find that bloom timing generally shifts later at mid-latitudes and earlier at high and low latitudes by ~5 days per decade to 2100. The spatial patterns of bloom timing are similar in both low (monthly) and high (5 day) resolution data, although initiation dates are later at low resolution. The magnitude of the trends in bloom timing from 2006 to 2100 is very similar at high and low resolution, with the result that the number of years of data needed to detect a trend in phytoplankton phenology is relatively insensitive to data temporal resolution. We also investigate the influence of spatial scales on bloom timing and find that trends are generally more rapidly detectable after spatial averaging of data. Our results suggest that, if pinpointing the start date of the spring bloom is the priority, the highest possible temporal resolution data should be used. However, if the priority is detecting long-term trends in bloom timing, data at a temporal resolution of 20 days are likely to be sufficient. Furthermore, our results suggest that data sources which allow for spatial averaging will promote more rapid trend detection.