IEDL-IDS: An Image-Enhanced Encoder-Based Deep Learning Scheme for Intrusion Detection Systems
Bibliographic record
Abstract
As networks expand and evolve, their increasing complexity introduces significant security challenges, necessitating robust Intrusion Detection Systems (IDS). Traditional IDS often struggle to detect sophisticated cyberattacks due to their reliance on raw network data and primitive feature extraction techniques. To address these limitations, we propose an Image-enhanced Encoder-based Deep Learning scheme for Intrusion Detection Systems (IEDL-IDS), which combines image-based transformation and encoder-based feature extraction to detect complex intrusion patterns in network traffic. Technically, IEDL-IDS consists of three sequential modules. The preprocessing module transforms raw network traffic into RGB images to reveal temporal and spatial patterns. Thereafter, the encoder module processes the RGB images to extract latent features. Finally, the classifier module utilizes the latent features for high-accuracy intrusion detection. Notably, IEDL-IDS is highly flexible, as its built-in classifier can be easily replaced with any neural network-based model. This feature highlights the adaptability of IEDL-IDS in balancing detection performance with resource constraints, thereby meeting the diverse needs of network security applications. Our experimental results demonstrate that IEDL-IDS outperforms the state-of-the-art IDS schemes. On the CICIoT dataset, IEDL-IDS achieves a classification accuracy of 99.91% for binary classification and 95.66% for multi-class classification. Similarly, it attains 99.61% and 98.25% accuracy on the NSL-KDD dataset, and 99.27% and 96.42% on the ToN_IoT dataset, for binary and multi-class tasks, respectively. Notably, despite its high detection performance, IEDL-IDS maintains a competitive computational footprint, making it a practical and scalable solution for real-world intrusion detection deployments.
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".