Deep learning has become a widely used powerful tool in many research fields, although not much so yet in agriculture technologies. In this work, two deep convolutional neural networks (CNN), viz. Residual Network (ResNet) and its improved version named ResNeXt, are used to detect internal mechanical damage of blueberries using hyperspectral transmittance data. The original structure and size of hypercubes are adapted for the deep CNN training. To ensure that the models are applicable to hypercube, we adjust the number of filters in the convolutional layers. Moreover, a total of 5 traditional machine learning algorithms, viz. Sequential Minimal Optimization (SMO), Linear Regression (LR), Random Forest (RF), Bagging and Multilayer Perceptron (MLP), are performed as the comparison experiments. In terms of model assessment, k-fold cross validation is used to indicate that the model performance does not vary with the different combination of dataset. In real-world application, selling damaged berries will lead to greater interest loss than discarding the sound ones. Thus, precision, recall, and F1-score are also used as the evaluation indicators alongside accuracy to quantify the false positive rate. The first three indicators are seldom used by investigators in the agricultural engineering domain. Furthermore, ROC curves and Precision-Recall curves are plotted to visualize the performance of classifiers. The fine-tuned ResNet/ResNeXt achieve average accuracy and F1-score of 0.8844/0.8784 and 0.8952/0.8905, respectively. Classifiers SMO/ LR/RF/Bagging/MLP obtain average accuracy and F1-score of 0.8082/0.7606/0.7314/0.7113/0.7827 and 0.8268/0.7796/0.7529/0.7339/0.7971, respectively. Two deep learning models achieve better classification performance than the traditional machine learning methods. Classification for each testing sample only takes 5.2 ms and 6.5 ms respectively for ResNet and ResNeXt, indicating that the deep learning framework has great potential for online fruit sorting. The results of this study demonstrate the potential of deep CNN application on analyzing the internal mechanical damage of fruit.

To achieve the simultaneous and unobtrusive breathing rate (BR) and heart rate (HR) measurements during nighttime, we leverage a far-infrared imager and an infrared camera equipped with IR-Cut lens and an infrared lighting array to develop a dual-camera imaging system. A custom-built cascade face classifier, containing the conventional Adaboost model and fully convolutional network trained by 32K images, was used to detect the face region in registered infrared images. The region of interest (ROI) inclusive of mouth and nose regions was afterwards confirmed by the discriminative regression and coordinate conversions of three selected landmarks. Subsequently, a tracking algorithm based on spatio-temporal context learning was applied for following the ROI in thermal video, and the raw signal was synchronously extracted. Finally, a custom-made time-domain signal analysis approach was developed for the determinations of BR and HR. A dual-mode sleep video database, including the videos obtained under environment where illumination intensity ranged from 0 to 3 Lux, was constructed to evaluate the effectiveness of the proposed system and algorithms. In linear regression analysis, the determination coefficient (R²) of 0.831 had been observed for the measured BR and reference BR, and this value was 0.933 for HR measurement. In addition, the Bland-Altman plots of BR and HR demonstrated that almost all the data points located within their own 95% limits of agreement. Consequently, the overall performance of the proposed technique is acceptable for BR and HR estimations during nighttime.

In the era of multimedia and Internet, the quick response (QR) code helps people obtain information from offline to online quickly. However, the QR code is often limited in many scenarios because of its random and dull appearance. Therefore, this article proposes a novel approach to embed hyperlinks into common images, making the hyperlinks invisible for human eyes but detectable for mobile devices equipped with a camera. Our approach is an end-to-end neural network with an encoder to hide messages and a decoder to extract messages. To maintain the hidden message resilient to cameras, we build a distortion network between the encoder and the decoder to augment the encoded images. The distortion network uses differentiable 3-D rendering operations, which can simulate the distortion introduced by camera imaging in both printing and display scenarios. To maintain the visual attraction of the image with hyperlinks, a loss function conforming to the human visual system (HVS) is used to supervise the training of the encoder. Experimental results show that the proposed approach outperforms the previous work on both robustness and quality. Based on the proposed approach, many applications become possible, for example, ``image hyperlinks'' for advertisement on TV, website, or poster, and ``invisible watermark'' for copyright protection on digital resources or product packagings.

Deepfake technology aims to synthesize highly realistic facial images and videos, with broad application potential in entertainment, film production, and digital human modeling. Deep learning has driven major progress in generative modeling, from VAEs and GANs to the recent rise of diffusion models. The latter have sparked a renewed wave of research through their superior generation quality. In addition to deepfake generation, corresponding detection technologies continuously evolve to regulate the potential misuse of deepfakes, such as privacy invasion and phishing attacks. This survey comprehensively reviews the latest developments in deepfake generation and detection, summarizing and analyzing current state-of-the-arts in this rapidly evolving field. First, we unify task definitions, comprehensively introduce datasets and metrics, and summarize the underlying technologies. Then, we review the development of several related sub-fields and examine four representative deepfake research fields: face swapping, face reenactment, talking-face generation, and facial attribute editing, as well as forgery detection. Subsequently, we benchmark representative methods on widely adopted datasets to provide a comprehensive and up-to-date evaluation of the most influential published works. Finally, we discuss the key challenges and outline future research directions for the field. We closely follow the latest developments in this project.