Dilated Strip-Wise Spatial Feature Pyramid: An Efficient Network for Object Detection
Bibliographic record
Abstract
Object detection has become a fundamental capability in modern computer vision, enabling critical applications from autonomous systems to aerial surveillance. Unmanned aerial vehicle object detection (UAV-OD) presents unique challenges, including small object sizes, occlusions, and complex back-grounds. While most existing approaches rely on standard feature pyramid networks (FPN) to combine multi-level features, these methods often fail to capture long-range contextual relationships and directional patterns essential for small object detection. To address these limitations, we propose dilated strip-wise spatial feature pyramid (DSSFP), a novel architecture that explicitly focuses on long-range dependencies through dilated strip-wise convolutions (Conv), directional features via spatial-aware attention mechanisms, and multi-scale context while preserving spatial resolution. On the VisDrone dataset, our method improves average precision (AP) by +22.3% and AP50 by +26.1% compared to the baseline. Our method establishes new state-of-the-art results on the AI-TOD benchmark (27.8% AP/59.9% AP50), improving the prior work by +3.3%. The consistent gains across small, medium, and large objects demonstrate the framework's robustness for UAV applications. Source code is available at: https://github.com/harish1120/HR-DSSFP
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".