The increasing consumption of ducks and chickens in China demands characterizing carcasses of domestic birds efficiently. Most existing methods, however, were developed for characterizing carcasses of pigs or cattle. Here, we developed a noncontact and automated weighing method for duck carcasses hanging on a production line. A 2D camera with its facilitating parts recorded the moving duck carcasses on the production line. To estimate the weight of carcasses, the images in the acquired dataset were modeled by a convolution neuron network (CNN). This model was trained and evaluated using 10-fold cross-validation. The model estimated the weight of duck carcasses precisely with a mean abstract deviation ( MAD ) of 58.8 grams and a mean relative error ( MRE ) of 2.15% in the testing dataset. Compared with 2 widely used methods, pixel area linear regression and the artificial neural network (ANN) model, our model decreases the estimation error MAD by 64.7 grams (52.4%) and 48.2 grams (45.0%). We release the dataset and code at https://github.com/RuoyuChen10/Image_weighing.

The collision between sea ice and ship is a serious threat to navigation safety. However, the idea of using vibration data generated from ship-ice collisions to obtain physical information about sea ice, such as its type and thickness, is a novel method. This method overcomes the problem of traditional techniques (optical cameras and satellite remote sensing) being affected by visibility, and also provides data for assessing the ice load capacity of ships. In this paper, we propose a new method for sea ice type and thickness identification based on vibration sensor networks and machine learning. Firstly, the characteristic data set of sea ice type and thickness is constructed by analyzing the six nodes ice-ship vibration signals obtained from the vibration sensor network. We then used six classification algorithms to identify three common sea ice types (broken ice, flat ice, sea/open water) and assessed the accuracy of these algorithms at each node. Then, we use Gradient Boosting Decision Tree (GBDT) regression algorithm to predict sea ice thickness and evaluate the combination of node features. The results shows that the accuracy of single-node sea ice classification is 98.71%, and the R ² for sea ice thickness regression on a single node is 0.908. After feature fusion of two nodes, the R ² for sea ice thickness regression is 0.94, with a Root Mean Square Error (RMSE) of 0.847 cm, and the accuracy is better than traditional identification methods. The results show that the sea ice type and thickness can be identified with high accuracy by analyzing the vibration signals generated by ice-ship collision. The reliability and effectiveness of the proposed method in low visibility and other environments are verified by the sailing experiments in the Bohai Sea ice area accumulated at 1020 nautical miles.

Vacuum packaging effectively prevents food contamination, facilitates portability and storage. Visual detection technology can rapidly identify packaging issues, ensure food quality, and is a crucial method to address food safety concerns. However, due to the extensive semantic interference and imbalanced distribution of defects in packaging images, defect detection faces significant challenges. In this article, a semantic-adaptive multi-task defect detection approach for vacuum packaging is proposed to handle these problems. First, a vacuum packaging dataset is created by capturing packaging samples on the production line using an industrial camera. Next, addressing the issue of semantic interference, a novel adaptive region partitioning method, integrating U-net and image processing techniques, is proposed to enhance the identification of defects across different regions. Finally, considering the imbalance in defect distribution, the Disentangled Representations of Abnormalities (DRA) model with diverse modes is employed for defect detection tasks across multiple regions in this paper. The introduction of a Convolutional Block Attention Module (CBAM) in the DRA network enhances the capability to detect authentic defects. Experimental results shows that the method achieves an AUC of 99.47% and an accuracy of 97.82% on the constructed dataset, showcasing significant improvement over alternative methods. In conclusion, this study not only enhances the defect detection capabilities of food vacuum packaging but also provides solutions to common issues such as semantic interference and imbalanced distribution of defects.

Sea ice poses a significant hazard to vessels navigating in high-latitude maritime regions. To effectively prevent the collision between ships and large sea ice (such as icebergs) and other similar accidents, it is crucial to identify sea ice and its types in a timely and accurate manner. However, traditional sea ice identification techniques, such as optical and radar remote sensing, are often limited by visibility conditions, suffering from poor accuracy, high measurement costs, and low identification efficiency. This paper proposes a sea ice identification method based on a vibration sensor network, combined with the PCA-BPMIV-GRU strategy, which achieves low cost, high efficiency, and high accuracy in sea ice type detection. Firstly, a cost-effective MEMS vibration sensor network is deployed to capture vibration information of sea ice. The data collected from six sensors at different locations at the same time are treated as independent samples to construct a dataset, which contains time-frequency domain characteristic data of three sea-ice types: broken ice, flat ice, and open water. Secondly, the dimension reduction and feature optimization of vibration data were carried out by combining Principal Component Analysis (PCA) and the improved Mean Influence Value Optimization algorithm (BPMIV). This integration of unsupervised and supervised algorithms effectively eliminated unnecessary feature information, significantly improving model identification efficiency. Finally, seven commonly used classification algorithms are compared for the identification of three types of sea ice, evaluating their identification accuracy and efficiency under different dimensionality reduction methods. The results confirm that the Gated Recurrent Unit (GRU) model achieves the best comprehensive performance. Experimental data shows that the proposed method reduces the number of features to 9, resulting in a 24.2 % decrease in parameters compared to the GRU model alone, while achieving a high identification accuracy of 93.2 %. The parameters and identification accuracy of the proposed method are superior to current sea ice identification technologies.

Artificial agents must adapt to non-stationary physical environments; however, contemporary deep reinforcement learning relies heavily on subjective scalar reward optimization and statistical curve-fitting. This renders systems fundamentally fragile, highly susceptible to shortcut learning, and prone to catastrophic forgetting. It remains unclear how robust causal abstraction and zero-shot generalization can objectively emerge without explicit algorithmic intervention. Here, we propose Cognitive Wave Dynamics (CWD), a physics-first framework that reframes intelligence as the objective filtering of reality through innate physical structures, rather than mere algorithmic computation. Continuous physical features (e.g., mechanical tension) are phase-encoded bounded by hard-limit receptor saturation-establishing an innate boundary of cognitive resolution. Subsequently, these phases diffract through an embodied multi-dimensional grating. By abandoning backpropagation, the system reconstructs causality through the law of wave energy deposition ( E ∝ A ² ), a process strictly driven by objective physical survival feedback. We demonstrate that this framework inherently shields agents from extreme scalar noise. Critically, when fundamental physical laws mutate, CWD drives a spontaneous thermodynamic phase transition to dismantle obsolete standing waves and rapidly recrystallize causality. Furthermore, by extracting underlying physical invariants through conjugate phase alignment, CWD achieves zero-shot metaphorical reasoning on visually entirely novel objects, effectively bypassing the shortcut collapse of deep learning. These results establish that robust embodied intelligence and zero-shot reasoning can emerge purely from wave mechanics and thermodynamic entropy minimization, offering a new objective, physics-constrained paradigm for artificial general intelligence.