CA3D: Convolutional-Attentional 3D Nets for Efficient Video Activity Recognition on the Edge

2025/05/26 by Gabriele Lagani, Fabrizio Falchi, Lagani, Gabriele +5
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #Context-Aware Activity Recognition Systems #FOS: Computer and information sciences #Human Pose and Action Recognition #Video Surveillance and Tracking Methods

paper · pdf · doi:10.48550/arxiv.2505.19928

openalex publication_date 2025/05/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we introduce a deep learning solution for video activity recognition that leverages an innovative combination of convolutional layers with a linear-complexity attention mechanism. Moreover, we introduce a novel quantization mechanism to further improve the efficiency of our model during both training and inference. Our model maintains a reduced computational cost, while preserving robust learning and generalization capabilities. Our approach addresses the issues related to the high computing requirements of current models, with the goal of achieving competitive accuracy on consumer and edge devices, enabling smart home and smart healthcare applications where efficiency and privacy issues are of concern. We experimentally validate our model on different established and publicly available video activity recognition benchmarks, improving accuracy over alternative models at a competitive computing cost.

Citations

• Synaptic Plasticity Models and Bio-Inspired Unsupervised Deep Learning: A Survey
• Spiking Neural Networks and Bio-Inspired Supervised Deep Learning: A Survey
• Rethinking Video ViTs: Sparse Video Tubes for Joint Image and Video Learning
• UniFormerV2: Spatiotemporal Learning by Arming Image ViTs with Video UniFormer
• UniFormer: Unified Transformer for Efficient Spatiotemporal Representation Learning
• Vision Transformer for Small-Size Datasets
• Highlighting the Importance of Reducing Research Bias and Carbon Emissions in CNNs
• VidTr: Video Transformer Without Convolutions
• ViViT: A Video Vision Transformer
• A Survey of Quantization Methods for Efficient Neural Network Inference
• An Image is Worth 16x16 Words, What is a Video Worth?
• Is Space-Time Attention All You Need for Video Understanding?
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Rethinking Attention with Performers
• Pooling Methods in Deep Neural Networks, a Review
• Linformer: Self-Attention with Linear Complexity
• X3D: Expanding Architectures for Efficient Video Recognition
• Generating Long Sequences with Sparse Transformers
• Low-bit Quantization of Neural Networks for Efficient Inference
• D3D: Distilled 3D Networks for Video Action Recognition
• SlowFast Networks for Video Recognition
• Multi-Fiber Networks for Video Recognition
• LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networks
• Quantizing deep convolutional networks for efficient inference: A whitepaper
• End-to-End Learning of Motion Representation for Video Understanding
• PACT: Parameterized Clipping Activation for Quantized Neural Networks
• Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference
• Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification
• A Closer Look at Spatiotemporal Convolutions for Action Recognition
• Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition
• Non-local Neural Networks
• Attention Is All You Need
• The Kinetics Human Action Video Dataset
• Hidden Two-Stream Convolutional Networks for Action Recognition
• ActionFlowNet: Learning Motion Representation for Action Recognition
• Layer Normalization
• Identity Mappings in Deep Residual Networks
• Deep Residual Learning for Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Two-Stream Convolutional Networks for Action Recognition in Videos
• UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

Related

• An IoT Based Framework For Activity Recognition Using Deep Learning Technique
• Impact of Three-Dimensional Video Scalability on Multi-View Activity Recognition using Deep Learning
• Human Activity Recognition for Edge Devices
• A Simple and Efficient Baseline for Video Action Recognition
• Human Activity Recognition Using Cascaded Dual Attention CNN and Bi-Directional GRU Framework