In recent years, the rapid prevalence of high-definition video in Internet of Things (IoT) systems has been directly facilitated by advances in imaging sensor technology. To adapt to limited uplink bandwidth, most media platforms opt to compress videos to bitrate streams for transmission. However, this compression often leads to significant texture loss and artifacts, which severely degrade the Quality of Experience (QoE). We propose a latent feature diffusion model (LFDM) for compressed video quality enhancement, which comprises a compact edge latent feature prior network (ELPN) and a conditional noise prediction network (CNPN). Specifically, we first pre-train ELPNet to construct a latent feature space that captures rich detail information for representing sharpness latent variables. Second, we incorporate these latent variables into the prediction network to iteratively guide the generation direction, thus resolving the problem that the direct application of diffusion models to temporal prediction disrupts inter-frame dependencies, thereby completing the modeling of temporal correlations. Lastly, we innovatively develop a Grouped Domain Fusion module that effectively addresses the challenges of diffusion distortion caused by naive cross-domain information fusion. Comparative experiments on the MFQEv2 benchmark validate our algorithm’s superior performance in terms of both objective and subjective metrics. By integrating with codecs and image sensors, our method can provide higher video quality.

The complexity of deep neural network models (DNNs) severely limits their application on devices with limited computing and storage resources. Knowledge distillation (KD) is an attractive model compression technology that can effectively alleviate this problem. Multi-teacher knowledge distillation (MKD) aims to leverage the valuable and diverse knowledge distilled by multiple teacher networks to improve the performance of the student network. Existing approaches typically rely on simple methods such as averaging the prediction logits or using sub-optimal weighting strategies to fuse distilled knowledge from multiple teachers. However, employing these techniques cannot fully reflect the importance of teachers and may even mislead student’s learning. To address this issue, we propose a novel Decoupled Multi-Teacher Knowledge Distillation based on Entropy (DE-MKD). DE-MKD decouples the vanilla knowledge distillation loss and assigns adaptive weights to each teacher to reflect its importance based on the entropy of their predictions. Furthermore, we extend the proposed approach to distill the intermediate features from multiple powerful but cumbersome teachers to improve the performance of the lightweight student network. Extensive experiments on the publicly available CIFAR-100 image classification benchmark dataset with various teacher-student network pairs demonstrated the effectiveness and flexibility of our approach. For instance, the VGG8|ShuffleNetV2 model trained by DE-MKD reached 75.25%|78.86% top-one accuracy when choosing VGG13|WRN40-2 as the teacher, setting new performance records. In addition, surprisingly, the distilled student model outperformed the teacher in both teacher-student network pairs.

This paper addresses the problem where optimizing a single production shop within a production chain may not improve the overall performance of the entire chain. To overcome this and synchronize the efficiency of each shop, methods are proposed to align the output rate of an upstream shop with the limited intake rate of its downstream shop. In the proposed methods, the output rate of the upstream shop is used to guide job scheduling, processing, and resource allocation in the shop. Simulation results from a real-world case study demonstrate that implementing this pull-based system reduces job earliness and tardiness by over 90% in the tested factory, where the upstream shop is a flexible job shop, leading to lower inventory costs, idling costs, and labor costs.