PDNet: Exploring Pruning and Distillation for Real-Time Worker Activity Recognition in Industrial Assembly

Malek Baba; Thinesh Thiyakesan Ponbagavathi; Alina Roitberg

cs.CVOct 19, 2026

PDNet: Exploring Pruning and Distillation for Real-Time Worker Activity Recognition in Industrial Assembly

Authors: Malek Baba, Thinesh Thiyakesan Ponbagavathi, Alina Roitberg

Organizations: Stuttgart University, Stuttgart, Germany · Institute for Artificial Intelligence (KI), Stuttgart University, Stuttgart, Germany · Intelligent Assistive Systems (IAS) group, University of Hildesheim, Hildesheim, Germany

Abstract

In this work, we address real-time fine-grained human activity recognition (HAR) in industrial assembly and introduce PDNet, a new framework that jointly exploits structured pruning and knowledge distillation for 3D convolutional networks. We start by training a 3D-CNN backbone in a supervised manner. Then, PDNet applies channel-wise structured pruning to obtain a more compact network. In the third step, our method is optimized using knowledge distillation from the large unpruned teacher to recover accuracy lost due to pruning. On the challenging MECCANO dataset with an X3D-M backbone, the pruned model without distillation reaches a Top-1 accuracy of 25.54 %, while PDNet raises this to 39.46 %, with only 6.90 MB model size and 3.78 GFLOPs. These results show that distillation within PDNet is highly effective in restoring pruned model performance and highlight its high potential for resource-constrained HAR in industrial assembly.

Figures & tables

Figure 1: Accuracy-latency-model size trade-off for fine-grained industrial activity recognition on the MECCANO dataset. The X-axis shows CPU inference time per clip (ms) and the Y-axis Top-1 accuracy, the circle area encodes model size. Our pruned X3D with knowledge distillation (PDNet, red) recovers most of the accuracy lost by the pruned baseline (purple).

Model	Top-1 test (%)	Top-5 test (%)
SlowFast	45.16	73.75
3D MobileNet	27.31	50.51
X3D-L	34.50	63.09
X3D-M	36.49	65.64

Table 1: Comparison of baseline models on the Meccano dataset against SlowFast Model

Ratio (%)	Size (MB)	Time (ms)	GFLOPs
0.0	21.70	107.72	6.20
5.0	10.75	98.32	5.65
10.0	9.71	83.75	5.14
15.0	8.73	83.33	4.62
20.0	7.78	81.17	4.18
25.0	6.90	80.77	3.78

Table 2: Pruning Results Summary

	Accuracy		Computational Efficiency
Model Variant	Top-1 (%)	Top-5 (%)	Size (MB)	Time (ms)	GFLOPs
Original X3D-M	36.49	65.64	21.70	107.72	6.20
Pruned Only	25.54	58.64	6.90	80.77	3.78
PDNet (Ours)	39.46	68.72	6.90	80.77	3.78

Table 3: Comparison of the different variants of the X3D-M model. The pruning and fine-tuning approach achieves large efficiency gains while maintaining competitive accuracy.

Human Activity Recognition

Human activity recognition classifies actions and behaviors from wearable sensors, video, radio signals, and other measurements.

See all 228 papers

Fast-HaMeR: Boosting Hand Mesh Reconstruction using Knowledge Distillation
Mar 17, 2026Hunain Ahmed Jillani, Ahmed Tawfik Aboukhadra, Ahmed Elhayek +33D Hand Reconstruction 3D Reconstruction
RPTU, Kaiserslautern, Germany · DFKI-AV, Kaiserslautern, Germany · UPM, Saudi Arabia +1
Label-Guided Knowledge Distillation for 3D-CNNs in Action Recognition
Sep 11, 2026Yanjiang Shi, Peng Zhao, Nan Qi +1Video Understanding Human Activity Recognition
School of Computer Science and Technology, and with the National Engineering Laboratory for Big Data Analytics (NEL-BDA), Xi’an Jiaotong University, Xi’an 710049, China
DANCE: Dynamic 3D CNN Pruning: Joint Frame, Channel, and Feature Adaptation for Energy Efficiency on the Edge
Mar 18, 2026Mohamed Mejri, Ashiqur Rasul, Abhijit ChatterjeeEnergy-Efficient ML Neural Network Pruning
School of Electrical and Computer Engineering, Georgia Tech, Atlanta, GA, USA

Mar 17, 2026J/K move · Enter open · S save

Fast-HaMeR: Boosting Hand Mesh Reconstruction using Knowledge Distillation

Hunain Ahmed Jillani, Ahmed Tawfik Aboukhadra, Ahmed Elhayek, Jameel Malik+2

RPTU, Kaiserslautern, Germany · DFKI-AV, Kaiserslautern, Germany · UPM, Saudi Arabia+1

3D Hand Reconstruction 3D Reconstruction Efficient Neural Network Inference Articulated Object Reconstruction Cross-Architecture Knowledge Distillation

Open paper AskPDF arXiv

Fast and accurate 3D hand reconstruction is essential for real-time applications in VR/AR, human-computer interaction, robotics, and healthcare. Most state-of-the-art methods rely on heavy models, limiting their use on resource-constrained devices like headsets, smartphones, and embedded systems. In this paper, we investigate how the use of lightweight neural networks, combined with Knowledge Distillation, can accelerate complex 3D hand reconstruction models by making them faster and lighter, while maintaining comparable reconstruction accuracy. While our approach is suited for various hand reconstruction frameworks, we focus primarily on boosting the HaMeR model, currently the leading method in terms of reconstruction accuracy. We replace its original ViT-H backbone with lighter alternatives, including MobileNet, MobileViT, ConvNeXt, and ResNet, and evaluate three knowledge distillation strategies: output-level, feature-level, and a hybrid of both. Our experiments show that using lightweight backbones that are only 35% the size of the original achieves 1.5x faster inference speed while preserving similar performance quality with only a minimal accuracy difference of 0.4mm. More specifically, we show how output-level distillation notably improves student performance, while feature-level distillation proves more effective for higher-capacity students. Overall, the findings pave the way for efficient real-world applications on low-power devices. The code and models are publicly available under https://github.com/hunainahmedj/Fast-HaMeR.

Abstract

Human Activity Recognition

Explore similar work

Fast-HaMeR: Boosting Hand Mesh Reconstruction using Knowledge Distillation

Label-Guided Knowledge Distillation for 3D-CNNs in Action Recognition

DANCE: Dynamic 3D CNN Pruning: Joint Frame, Channel, and Feature Adaptation for Energy Efficiency on the Edge