Multimodal Machine Learning for Pedestrian Detection
Bibliographic record
Abstract
Designing and developing autonomous vehicles that are capable of moving safely on roads by sensing the environment has motivated researchers to focus on pedestrian detection systems so they can detect people as fast and accurately as possible. However, for pedestrian detection, it is crucial to consider not only the pedestrians themselves but their color as well, because color has the advantage of being invariant to changes in scaling, rotation, and partial occlusion. Therefore, considering skin color detection for implementing pedestrian detection systems is an essential required step in ensuring autonomous vehicles are further incorporated into our society. Detecting human skin has proven to be a challenging problem because skin color can vary dramatically in its appearance due to many factors such as illumination, race, imaging conditions, and others. Recently it has been noted that pedestrian detection systems for autonomous vehicles perform poorly at detecting people with darker skin tones. Such findings indicate that there is a larger problem that is causing these issues: algorithmic bias. Algorithmic bias in pedestrian detection systems could be the leading factor of their poor performance due to the methods implemented and datasets used. Unfortunately, the algorithmic bias in this context has not been considered closely and it seems that many studies do not cover this aspect closely when discussing pedestrian detection systems for autonomous vehicles. To alleviate this, we attempt to explore different techniques that can be used to detect pedestrians while minimizing bias. In our experiment, we use both a YOLO v3 convolutional neural network and K-Means clustering for classifying skin-tones.
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 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| 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) | 0.006 | 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".