Lightweight Deep Learning for Visual Perception: A Survey of Models, Compression Strategies, and Edge Deployment Challenges
Bibliographic record
Abstract
The increasing demand for the deployment of deep neural networks (DNNs) in edge devices has led to the development of lightweight deep learning (LDL) models designed to operate efficiently under resource constraints. Although DNNs have achieved remarkable success in various applications, their high computational requirements often limit their deployment on devices with restricted memory and processing power. This challenge has motivated researchers to develop optimized LDL models that balance accuracy, speed, and efficiency while maintaining competitive performance. Despite existing surveys covering specific aspects of LDL models, a comprehensive review encompassing image classification, object detection, and segmentation remains limited. This proposed survey systematically explores recent advancements in LDL models, highlighting their architectures, optimization techniques, and real-world applications. This survey conducts an empirical evaluation by testing latest state-of-the-art LDL models on the Jetson Orin edge device using benchmark datasets: Caltech-256 for classification, VisDrone for object detection, and COCO for segmentation. The experimental analysis focuses on key performance metrics, including inference speed, model accuracy, and computational efficiency, while comparing LDL models with their conventional counterparts. This study provides a holistic understanding of the role of LDL models in edge computing, providing insight into emerging trends, challenges, and future research directions in the field.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".