Overview
Who this competition fits
适合全日制普通高等教育在校学生,且有目标检测与 PaddlePaddle 模型部署经验的人;线上可个人或最多 5 人组队,但代表学校进入线下分区赛的队伍名额有限。尤其适合希望围绕农场场景的 20 类目标识别,完成可运行推理交付物的学生团队。
Read the original official blurb
第二十一届全国大学生智能汽车竞赛——百度智慧交通组创意赛 Official tags: OCR+Det+文心. Sign-ups: 332.
Resources
What this competition gives you
Taken from the competition's official page.
Preparation
From registration to a first submission
- 01
PaddlePaddle 推理部署
- 02
目标检测
- 03
F1-score 与 IoU
- 04
Python 3.10
- 05
JSON 结果生成
- 06
GPU 推理优化
Before you commit: 核心难点是同时满足检测精度和端到端速度约束:前处理、推理、后处理合计必须达到 20 FPS,否则直接记 0 分。还需把模型、推理脚本和依赖整理成指定的 submission.zip,在固定 PaddlePaddle、Python 3.10 与 V100 环境中正确生成 result.json。
Source
How this page was assembled
Competition information is structured from the official page. Scores are platform estimates for decision support; official rules take precedence.
- Official competition page
- Baidu AI Studio
- Last checked
- Not recorded