File Fragment Type Classification Using Light-Weight Convolutional Neural Networks
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
In digital forensics, file carving is used to extract files without relying on the underlying file system metadata. This process can be challenging if the file is fragmented. Therefore, it is important first to identify the type of file fragment. There exist several techniques to identify the type of file fragments without relying on metadata, for example, using headers and footers to identify the fragment type. Recently, convolutional neural network (CNN) models have been used to build classification models to achieve this task. Existing models for file fragment type classification often require significant computational resources due to their large number of parameters, leading to slower inference times and higher memory consumption. To address these challenges, we propose light-weight file fragment type classification models based on separable CNNs that maintain comparable accuracy while reducing computational demands. Our proposed light-weight file fragment type classification model leverages depthwise separable convolutions to improve the efficiency of feature extraction while reducing computational overhead. This approach leads to improved classification performance by focusing on the most relevant features within file fragments, achieving comparable accuracy to state-of-the-art models with significantly fewer parameters. The evaluation results demonstrate the model’s effectiveness, with a 79% accuracy on the FFT-75 dataset using nearly 100K parameters and 164M FLOPs —representing a 4x reduction in model size and a 6x improvement in speed over the best-performing existing classifier. Our results demonstrate that these light-weight models are effective for real-time digital forensic applications where computational efficiency is critical.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it