Runtime-dependent rankings of detector design choices: code, measurements, trained weights and run logs
Year: 2026
Type: dataset
Abstract: Code, raw measurements, trained weights and run logs for the paper “Does the Deployment Runtime Change Which Detector Design Wins? Host-Bound and Device-Bound Latency Regimes of YOLO11-Based Detectors for UAV Imagery” (submitted to the Journal of Real-Time Image Processing). runtime-dependent-rankings-main.zip: snapshot of https://github.com/langyuan8080-ai/runtime-dependent-rankings (measurement and analysis scripts, all latency, profiling and accuracy records; bash reproduce.sh recomputes every table and figure of the paper). weights_best_68.tar: best.pt weights of the 68 YOLO11-based training runs on VisDrone2019-DET used in the paper (layout runs/detect/<run>/weights/best.pt; checksums in SHA256SUMS_weights.txt). run_logs.tar.gz: logs of all measurement runs. The code is licensed under AGPL-3.0, the data and weights under CC BY 4.0.
Authors: Lang, LinFang, Kim, Jin-whan
Institution: Youngsan University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7224929308