Image Enhancement for Better VRU Detection in Challenging Weather Conditions
Bibliographic record
Abstract
This paper introduces a new approach to improve the performance of object detection techniques in challenging weather conditions through image enhancement. In 2018, a tragic accident occurred while testing a self-driving vehicle under severe bad lighting conditions, where a person crossing the street with a bicycle was not detected early enough for the vehicle to take action, which led to the death of the cyclist. This shows the dire need for robust, fast, and efficient detection techniques in adversarial conditions. In this paper, we propose a novel processing pipeline that contains a multilabel classifier trained to capture quality issues in a given frame and then passes it to proper filters to enhance the quality. The image qualities we target in this study include low visibility/bad lighting conditions, blurring, and image distortion. The raw frame goes through a multi-label classifier first to detect possible quality degradation. The frame then passes through various filters to fix the detected qualities accordingly. Performance evaluation shows that the proposed multilabel classifier achieves 80% accuracy during testing with a 20-ms average inference time for each frame. The proposed pipeline also enhances the frame processing delay by 3x and increases the confidence score of the YOLO detection algorithm by 5% on average.
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 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".