With the rise of artificial intelligence, deep learning is gradually applied to the field of agriculture and plant science. However, the excellent performance of deep learning needs to be established on massive numbers of samples. In the field of plant science and biology, it is not easy to obtain a large amount of labeled data. The emergence of few-shot learning solves this problem. It imitates the ability of humans’ rapid learning and can learn a new task with only a small number of labeled samples, which greatly reduces the time cost and financial resources. At present, the advanced few-shot learning methods are mainly divided into four categories based on: data augmentation, metric learning, external memory, and parameter optimization, solving the over-fitting problem from different viewpoints. This review comprehensively expounds on few-shot learning in smart agriculture, introduces the definition of few-shot learning, four kinds of learning methods, the publicly available datasets for few-shot learning, various applications in smart agriculture, and the challenges in smart agriculture in future development.

Background: Learning from a few samples to automatically recognize the plant leaf diseases is an attractive and promising study to protect the agricultural yield and quality. The existing few-shot classification studies in agriculture are mainly based on supervised learning schemes, ignoring unlabeled data's helpful information. Methods: In this paper, we proposed a semi-supervised few-shot learning approach to solve the plant leaf diseases recognition. Specifically, the public PlantVillage dataset is used and split into the source domain and target domain. Extensive comparison experiments considering the domain split and few-shot parameters (N-way, k-shot) were carried out to validate the correctness and generalization of proposed semi-supervised few-shot methods. In terms of selecting pseudo-labeled samples in the semi-supervised process, we adopted the confidence interval to determine the number of unlabeled samples for pseudo-labelling adaptively. Results: The average improvement by the single semi-supervised method is 2.8%, and that by the iterative semi-supervised method is 4.6%. Conclusions: The proposed methods can outperform other related works with fewer labeled training data.

The path planning of the autonomous underwater vehicle (AUV) has shown great potential in various Internet of Underwater Things (IoUT) applications. Although considerable efforts had been made, prior studies are confronted with some limitations. For one thing, existing work only uses the ocean current simulation model without introducing real ocean information, having not been supported by real data. For another, traditional path planning algorithms have strong environment dependence and lack flexibility: once the environment changes, they need to be remodeled and replanned. To overcome these challenges, this article proposes comprehensive ocean information D3QN (COID), an AUV path planning scheme exploiting comprehensive ocean information and reinforcement learning (RL), which consists of three steps. First, we introduce the comprehensive real ocean data, including weather, temperature, thermohaline, current, etc., and apply them into the regional ocean modeling system to generated reliable ocean current. Next, through well-designed state transition function and reward function, we build a 3-D grid model of ocean environment for RL. Furthermore, based on the framework of the double dueling deep $Q$ network (D3QN), COID integrates local ocean current and position features to provide state input and uses priority sampling to accelerate network convergence. The performance of COID has been evaluated and proved by numerical results, which demonstrate efficient path planning and high flexibility for expansion into different ocean environments.

As one of the commonly used vehicles for underwater detection, underwater robots are facing a series of problems. The real underwater environment is large-scale, complex, real-time and dynamic, and many unknown obstacles may exist in the underwater environment. Under such complex conditions and lack of prior knowledge, the existing path planning methods are difficult to plan, therefore they cannot effectively meet the actual demands. In response to these problems, a three-dimensional marine environment including multiple obstacles is established with the real ocean current data in this paper, which is consistent with the actual application scenarios. Then, we propose an N-step Priority Double DQN (NPDDQN) path planning algorithm, which potently realizes obstacle avoidance in the complex environment. In addition, this study proposes an experience screening mechanism, which screens the explored positive experience and improves its reuse rate, thus efficiently improving the algorithm stability in the dynamic environment. This paper verifies the better performance of reinforcement learning compared with a variety of traditional methods in three-dimensional underwater path planning. Underwater robots based on the proposed method have good autonomy and stability, which provides a new method for path planning of underwater robots. Note to Practitioners—The goal of this study is to provide a new solution for obstacle avoidance in path planning of underwater robots, which is consistent with the dynamic and real-time demands of the real environment. Existing underwater path planning researches lack a consistent environment with the actual application, and therefore we firstly construct a three-dimensional ocean environment with real ocean current data to provide support for the algorithms. Additionally, most of the algorithms are pre-planning methods or require long-time calculation, and there is little research on obstacle avoidance. In the face of obstacle changes, underwater robots with poor adaptability will cause performance decline and even economic losses. The proposed algorithm learns through interaction with the environment, and therefore it does not require any prior experience, and has good adaptability as well as fast inference speed. Especially, in the dynamic environment, algorithm performance is difficult to guarantee due to less positive experience in exploration. The proposed experience screening mechanism improves the stability of the algorithm, so that the underwater robot maintains stable performance in different dynamic environments.