Early blight and late blight are important factors restricting tomato yield. However, it is still a challenge to accurately and objectively detect and segment crop diseases in order to evaluate disease damage. In this paper, the Disease Segmentation Detection Transformer (DS-DETR) is proposed to segment leaf disease spots efficiently based on several improvements to DETR. Additionally, a damage assessment is carried out by the area ratio of the segmented leaves to the disease spots. First, an unsupervised pre-training method was introduced into DETR with the Plant Disease Classification Dataset (PDCD) to solve the problem of the long training epochs and slow convergence speed of DETR. This method can train the Transformer structures in advance to obtain leaf disease features. Loading the pre-training model weight in DS-DETR can speed up the convergence speed of the model. Then, Spatially Modulated Co-Attention (SMCA) was used to assign Gaussian-like spatial weights to the query box of DS-DETR. The different positions in the image are trained using the query boxes with different weights to improve the accuracy of the model. Finally, an improved relative position code was added to the Transformer structure of DS-DETR. Relative position coding promotes the capture of the sequence order of input tokens by the Transformer. The spatial location feature is strengthened by establishing the location relationship between different instances. Based on these improvements, the DS-DETR model was tested on the Tomato leaf Disease Segmentation Dataset (TDSD) constructed by us. The experimental results show that the DS-DETR proposed by us achieved 0.6823 for AP_mask, which improved by 12.87%, 8.25%, 3.67%, 1.95%, 10.27%, and 9.52% compared with the state-of-the-art: Mask RCNN, BlendMask, CondInst, SOLOv2, ISTR, and DETR, respectively. In addition, the disease grading accuracy reached 0.9640 according to the segmentation results given by our proposed model.

In complex citrus orchard environments, light changes, branch shading, and fruit overlapping impact citrus detection accuracy. This paper proposes the citrus detection model YOLO-DCA in complex environments based on the YOLOv7-tiny model. We used depth-separable convolution (DWConv) to replace the ordinary convolution in ELAN, which reduces the number of parameters of the model; we embedded coordinate attention (CA) into the convolution to make it a coordinate attention convolution (CAConv) to replace the ordinary convolution of the neck network convolution; and we used a dynamic detection head to replace the original detection head. We trained and evaluated the test model using a homemade citrus dataset. The model size is 4.5 MB, the number of parameters is 2.1 M, mAP is 96.98%, and the detection time of a single image is 5.9 ms, which is higher than in similar models. In the application test, it has a better detection effect on citrus in occlusion, light transformation, and motion change scenes. The model has the advantages of high detection accuracy, small model space occupation, easy application deployment, and strong robustness, which can help citrus-picking robots and improve their intelligence level.

Video action recognition faces persistent challenges in balancing accuracy with computational efficiency. While state space models, such as Mamba, have emerged with linear complexity advantages, they exhibit inefficiency in capturing critical spatiotemporal dependencies within video data. To address this core limitation, this paper proposes ETMamba, an enhanced architecture built upon the Mamba baseline. The ETMamba achieve performance breakthroughs via three core innovation modules: (1) the Spatiotemporal Feature Preservation module retains complete original spatiotemporal correlations before data flattening, solving the problem of spatiotemporal feature loss; (2) the Efficient Bidirectional Sharing strategy accurately models bidirectional temporal dependencies, enhancing key temporal dynamic information; and (3) the Spatiotemporal Collaborative Modulation mechanism combines global temporal and local spatial information to achieve collaborative capture of long-short term dependencies and fine-grained features. We conduct experiments on multiple benchmark datasets, achieving recognition accuracies of 88.3%, 74.6%, 75.7%, and 98.1% on Kinetics-400, Something-Something V2, HMDB-51, and Breakfast datasets, respectively, while maintaining low to medium computational complexity.