Haidong Kang

Dr. Haidong Kang


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 February 2024 in IEEE Transactions on Computers
Verified
Journal Article
Closed access
LMLianbo Ma
Lianbo Ma
HKHaidong Kang
Haidong Kang
GYGuo Yu
Guo Yu
QLQing Li
Qing Li
WWWenliang Wu
Wenliang Wu
Performance predictors are used to reduce architecture evaluation costs in neural architecture search, which however suffers from a large amount of budget consumption in annotating substantial architectures trained from scratch. Hence, how to leverage existing annotated architectures to train a generalized predictor to find the optimal architecture on unseen target search spaces becomes a new research topic. To solve this issue, we propose a Single-Domain Generalized Predictor (SDGP), which aims to make the predictor only trained on a single source search space but perform well on target search spaces. In meta-learning, we firstly adopt feature extractor in learning the domain-invariant features of the architectures. Then, a neural predictor is trained to map the architectures to the accuracy of the candidate architectures over the target domain simulated on the source search space. Moreover, a novel multi-head attention driven regularizer is designed to regulate the predictor to further improve the generalization ability of the predictor for the feature extractor. A series of experimental results have shown that the proposed predictor outperforms the state-of-the-art predictors in generalization and achieves significant performance gains in finding the optimal architectures with test error 2.40% on CIFAR-10 and 23.20% on ImageNet1k within 0.01 GPU days.

14 November 2025 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
JCJian Cheng
Jian Cheng
HKHaidong Kang
Haidong Kang
YSYuxin Shao
Yuxin Shao
NLNan Li
Nan Li
PCPengjun Chen
Pengjun Chen
LMLianbo Ma
Lianbo Ma
With the rapid advancement of large language models (LLMs) in both academia and industry, their growing size and complexity have introduced significant challenges in terms of computational cost and deployment efficiency. To address these issues, a wide range of inference optimization techniques—including but not limited to model compression—have been proposed to accelerate LLM inference while preserving model performance. This survey provides a comprehensive overview of LLM inference acceleration strategies, analyzing them from multiple perspectives, including foundational principles, algorithmic techniques, real-world applications, and open research challenges. We begin by introducing core concepts underlying inference optimization and propose a new taxonomy that categorizes existing approaches, including quantization, pruning, distillation, efficient architectures, compilation, and hardware-aware methods. Following the lifecycle of LLM development and deployment, we examine how these techniques interact with model training, fine-tuning, and serving. Furthermore, we highlight key applications of efficient LLMs and discuss emerging trends and unresolved issues in the field. By synthesizing recent advances, this survey aims to provide actionable insights and practical guidance for researchers and practitioners working with scalable and efficient LLM systems.

5 August 2024 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
HWHanling Wang
Hanling Wang
QLQing Li
Qing Li
HKHaidong Kang
Haidong Kang
DHDieli Hu
Dieli Hu
LMLianbo Ma
Lianbo Ma
YJYong Jiang
Yong Jiang
Real-time video analytics on edge devices has gained increasing attention across a wide range of business areas. However, edge devices usually have limited computing resources. Consequently, conventional approaches to video analytics either deploy simplified models on the edge (resulting in low accuracy) or transmit video content to the cloud (resulting in high latency and network overheads) to enable deep learning inference (e.g., object detection). In this paper, we introduce ParaLoupe, a novel real-time video analytics system that parallelizes deep learning inference in the edge cluster with task-oriented mini models. These mini models do not attain State-of-the-Art accuracy individually, but collectively can achieve much better accuracy-latency tradeoff than State-of-the-Art models. To achieve this, ParaLoupe crops multiple single-object patches from a given video frame. These single-object patches are then sent to multiple edge devices for parallel inference with specifically designed mini models. A patch-based task scheduling algorithm is further proposed to leverage the computing resources of the edge cluster to meet the service-level objectives. Our experimental results on real-world datasets show that ParaLoupe significantly outperforms baseline methods, achieving up to 14.1× inference speedup with accuracy on par with state-of-the-art models, or improving accuracy up to 45.1% under the same latency constraints.

14 February 2025 in Mathematics
Verified
Journal Article
Open access
PDF file
JCJian Cheng
Jian Cheng
JJJinbo Jiang
Jinbo Jiang
HKHaidong Kang
Haidong Kang
LMLianbo Ma
Lianbo Ma
Coal mine scene image recognition plays an important role in safety monitoring and equipment detection. However, traditional methods often depend on manually designed neural network architectures. These models struggle to handle the complex backgrounds, low illumination, and diverse objects commonly found in coal mine environments. Manual designs are not only inefficient but also restrict the exploration of optimal architectures, resulting to subpar performance. To address these challenges, we propose using a neural architecture search (NAS) to automate the design of neural networks. Traditional NAS methods are known to be computationally expensive. To improve this, we enhance the process by incorporating Particle Swarm Optimization (PSO), a scalable algorithm that effectively balances global and local searches. To further enhance PSO’s efficiency, we integrate the lifespan mechanism, which prevents premature convergence and enables a more comprehensive exploration of the search space. Our proposed method establishes a flexible search space that includes various types of convolutional layers, activation functions, pooling operations, and network depths, enabling a comprehensive optimization process. Extensive experiments show that the Lifespan-PSO NAS method outperforms traditional manually designed networks and standard PSO-based NAS approaches, offering significant improvements in both recognition accuracy (improved by 10%) and computational efficiency (resource usage reduced by 30%). This makes it a highly effective solution for real-world coal mine image recognition tasks via a PSO-optimized approach in terms of performance and efficiency.


1 September 2025
Verified
Conference Paper
Open access
PDF file
HKHaidong Kang
Haidong Kang
LMLianbo Ma
Lianbo Ma
GYGuo Yu
Guo Yu
Shangce GaoShangce Gao
Shangce Gao
Mixed precision quantization (MPQ) is an effective quantization approach to achieve accuracy-complexity trade-off of neural network, through assigning different bit-widths to network activations and weights in each layer. The typical way of existing MPQ methods is to optimize quantization policies (i.e., bit-width allocation) in a gradient descent manner, termed as Differentiable MPQ (DMPQ). At the end of the search, the bit-width associated to the quantization parameters which has the largest value will be selected to form the final mixed precision quantization policy, with the implicit assumption that the values of quantization parameters reflect the operation contribution to the accuracy improvement. While much has been discussed about the MPQ’s improvement, the bit-width selection process has received little attention. We study this problem and argue that the magnitude of quantization parameters does not necessarily reflect the actual contribution of the bit-width to the task performance. Then, we propose a Shapley-based MPQ (SMPQ) method, which measures the bit-width operation’s direct contribution on the MPQ task. To reduce computation cost, a Monte Carlo sampling-based approximation strategy is proposed for Shapley computation. Extensive experiments on mainstream benchmarks demonstrate that our SMPQ consistently achieves state-of-the-art performance than gradient-based competitors.