Convolutional neural networks for single-image super-resolution have been widely used with great success. However, most of these methods use L1 loss to guide network optimization, resulting in blurry restored images with sharp edges smoothed. This is because L1 loss limits the optimization goal of the network to the statistical average of all solutions within the solution space of that task. To go beyond the L1 loss, this paper designs an image super-resolution algorithm based on second-order gradient loss. We impose additional constraints by considering the high-order gradient level of the image so that the network can focus on the recovery of fine details such as texture during the learning process. This helps to alleviate the problem of restored image texture over-smoothing to some extent. During network training, we extract the second-order gradient map of the generated image and the target image of the network by minimizing the distance between them, this guides the network to pay attention to the high-frequency detail information in the image and generate a high-resolution image with clearer edge and texture. Besides, the proposed loss function has good embeddability and can be easily integrated with existing image super-resolution networks. Experimental results show that the second-order gradient loss can significantly improve both Learned Perceptual Image Patch Similarity (LPIPS) and Frechet Inception Distance score (FID) performance over other image super-resolution deep learning models.

There is a very significant knowledge gap between Artificial Intelligence (AI) and a multitude of industries that exist in today’s modern world. This is primarily attributable to the limited availability of resources and technical expertise. However, a major obstacle is that AI needs to be flexible enough to work in many different applications, utilising a wide variety of datasets through cloud computing. As a result, we developed a benchmark toolkit called CloudAISim to make use of the power of AI and cloud computing in order to satisfy the requirements of modern applications. The goal of this study is to come up with a strategy for building a bridge so that AI can be utilised in order to assist those who are not very knowledgeable about technological advancements. In addition, we modelled a healthcare application as a case study in order to verify the scientific reliability of the CloudAISim toolkit and simulated it in a cloud computing environment using Google Cloud Functions to increase its real-time efficiency. A non-expert-friendly interface built with an interactive web app has also been developed. Any user without any technical knowledge can operate the entire model, which has a 98% accuracy rate. The proposed use case is designed to put AI to work in the healthcare industry, but CloudAISim would be useful and adaptable for other applications in the future.

As neural network inference demands are ever-increasing in intelligent applications, the performance optimization of model serving becomes a challenging problem. Dynamic batching is an important feature of contemporary deep learning serving systems, which combines multiple requests of model inference and executes them together to improve the system’s throughput. However, the behavior characteristics of each part in deep neural network batching systems as well as their performance impact on different model structures are still unknown. In this paper, we characterize the batching system by leveraging three representative deep neural networks on GPUs, performing a systematic analysis of the performance effects from the request batching module, model slicing module, and stage reorchestrating module. Based on experimental results, several insights and recommendations are offered to facilitate the system design and optimization for deep learning serving.

The burgeoning field of Artificial Intelligence Generated Content (AIGC) is witnessing rapid advancements, particularly in video generation. This paper introduces AIGCBench, a pioneering comprehensive and scalable benchmark designed to evaluate a variety of video generation tasks, with a primary focus on Image-to-Video (I2V) generation. AIGCBench tackles the limitations of existing benchmarks, which suffer from a lack of diverse datasets, by including a varied and open-domain image–text dataset that evaluates different state-of-the-art algorithms under equivalent conditions. We employ a novel text combiner and GPT-4 to create rich text prompts, which are then used to generate images via advanced Text-to-Image models. To establish a unified evaluation framework for video generation tasks, our benchmark includes 11 metrics spanning four dimensions to assess algorithm performance. These dimensions are control-video alignment, motion effects, temporal consistency, and video quality. These metrics are both reference video-based and video-free, ensuring a comprehensive evaluation strategy. The evaluation standard proposed correlates well with human judgment, providing insights into the strengths and weaknesses of current I2V algorithms. The findings from our extensive experiments aim to stimulate further research and development in the I2V field. AIGCBench represents a significant step toward creating standardized benchmarks for the broader AIGC landscape, proposing an adaptable and equitable framework for future assessments of video generation tasks. We have open-sourced the dataset and evaluation code on the project website: https://www.benchcouncil.org/AIGCBench .

This paper explores the paradigm of leveraging ChatGPT as a benchmark tool for theory prototyping in conceptual research. Specifically, we conducted two experimental studies using the classical technology acceptance model (TAM) to demonstrate and evaluate ChatGPT's capability of comprehending theoretical concepts, discriminating between constructs, and generating meaningful responses. Results of the two studies indicate that ChatGPT can generate responses aligned with the TAM theory and constructs. Key metrics including the factors loading, internal consistency reliability, and convergence reliability of the measurement model surpass the minimum threshold, thus confirming the validity of TAM constructs. Moreover, supported hypotheses provide an evidence for the nomological validity of TAM constructs. However, both of the two studies show a high Heterotrait–Monotrait ratio of correlations (HTMT) among TAM constructs, suggesting a concern about discriminant validity. Furthermore, high duplicated response rates were identified and potential biases regarding gender, usage experiences, perceived usefulness, and behavioural intention were revealed in ChatGPT-generated samples. Therefore, it calls for additional efforts in LLM to address performance metrics related to duplicated responses, the strength of discriminant validity, the impact of prompt design, and the generalizability of findings across contexts.