Abstract:Object detection based on Knowledge Distillation can enhance the capabilities and performance of 5G and 6G networks in various domains, such as autonomous vehicles, smart surveillance, and augmented reality. The integration of object detection with Knowledge Distillation techniques is expected to play a pivotal role in realizing the full potential of these networks. This study presents Shared Knowledge Distillation (Shared-KD) as a solution to overcome optimization challenges caused by disparities in cross-layer features between teacher–student networks. The significant gaps in intermediate-level features between teachers and students present a considerable obstacle to the efficacy of distillation. To tackle this issue, we draw inspiration from collaborative learning in real-world education, where teachers work together to prepare lessons and students engage in peer learning. Building upon this concept, our innovative contributions in model construction are highlighted as follows: (1) A teacher knowledge augmentation module: this module is proposed to combine lower-level teacher features, facilitating the knowledge transfer from the teacher to the student. (2) A student mutual learning module is introduced to enable students to learn from each other, mimicking the peer learning concept in collaborative learning. (3) The Teacher Share Module combines lower-level teacher features: the specific functionality of the teacher knowledge augmentation module is described, which involves combining lower-level teacher features. (4) The multi-step transfer process can be easily optimized due to the minimal gap between the features: the proposed approach breaks down the knowledge transfer process into multiple steps, which can be easily optimized due to the minimal gap between the features involved in each step. Shared-KD uses simple feature losses without additional weights in transformation, resulting in an efficient distillation process that can be easily combined with other methods for further improvement. The effectiveness of our approach is validated through experiments on popular tasks such as object detection and instance segmentation.

Distillation knowledge-based space-time data prediction on industrial IoT edge devices

Research on Knowledge Distillation Algorithm of Object Detection

Knowledge Distillation on Spatial-Temporal Graph Convolutional Network for Traffic Prediction

KDnet-RUL: A Knowledge Distillation Framework to Compress Deep Neural Networks for Machine Remaining Useful Life Prediction

Learning Temporal and Spatial Features Jointly: A Unified Framework for Space-Time Data Prediction in Industrial IoT networks

Knowledge Distillation via Token-Level Relationship Graph Based on the Big Data Technologies

Generative Denoise Distillation: Simple Stochastic Noises Induce Efficient Knowledge Transfer for Dense Prediction

KDGAN: Knowledge Distillation with Generative Adversarial Networks

Distilling Knowledge with Adversarial Networks

Directed Acyclic Graph Factorization Machines for CTR Prediction via Knowledge Distillation

Channel-wise Knowledge Distillation for Dense Prediction

Knowledge Distillation with a Precise Teacher and Prediction with Abstention

Distillation Sparsity Training Algorithm for Accelerating Convolutional Neural Networks in Embedded Systems

EdgeKE: An On-Demand Deep Learning IoT System for Cognitive Big Data on Industrial Edge Devices

Shared Knowledge Distillation Network for Object Detection

Highlight Every Step: Knowledge Distillation via Collaborative Teaching

Small Scale Data-Free Knowledge Distillation

A Lightweight and Adaptive Knowledge Distillation Framework for Remaining Useful Life Prediction

Ensemble Knowledge Distillation for CTR Prediction

Data Efficient Stagewise Knowledge Distillation

Improving Knowledge Distillation With a Customized Teacher