Benchmarking NVFlare Federated Algorithms in Decentralized Parking Space Detection and Classification Framework
Bibliographic record
Abstract
Most existing image-based parking space detection and classification methods assume that all training data reside in a single, centralized location—an unrealistic scenario that yields models unable to generalize to new parking lots. Furthermore, privacy concerns prevent lot owners from sharing raw images, limiting collaboration. To overcome these challenges, we present ParkFL, the first federated learning framework for parking space detection and classification that trains models across distributed client sites without exchanging raw image data. Built on NVFlare, ParkFL demonstrates model-agnosticism through evaluation on two deep-learning architectures. We benchmark four federated algorithms—FedAvg, FedProx, FedOpt, and SCAFFOLD—using real-world datasets. Despite training on non-centralized, heterogeneous data, ParkFL achieves 99.5% mAP, matching the accuracy of state-of-the-art centralized models on the same parking lots. When evaluated on images from different parking lots, ParkFL significantly outperforms models trained solely on individual-site data, which achieve 25.6% mAP, even though ParkFL never accesses raw images from other sites. Communication overhead remains below 3% of total training time, demonstrating a scalable, privacy-preserving solution with nearly state-of-the-art performance. We release ParkFL code at https://github.com/ahmedmbakr/ParkFL.
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 | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.003 | 0.002 |
| 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".