从coco网站运行test-dev2017图像以检测边界框对象,并将其提交到test-dev2018服务器。
收到的错误:
/opt/conda/lib/python2.7/site-packages/matplotlib/font_manager.py:273: UserWarning: Matplotlib is building the font cache using fc-list. This may take a moment.
warnings.warn('Matplotlib is building the font cache using fc-list. This may take a moment.')
Traceback (most recent call last):
File "/tmp/codalab/tmpJPSweB/run/program/run.py", line 111, in
res = cut_res(res, maxDet=100)
File "/tmp/codalab/tmpJPSweB/run/program/run.py", line 75, in cut_res
r_img = res[start_end[image_id][0]:start_end[image_id][1]]
KeyError: 581918
提交的.json文件只有6MB,zip文件只有900KB。当我手动将.json文件拆分为较小的.json(每个文件约1MB)时,可以成功评估1MB .json文件。我能否仅评估每个1MB .json文件并将每次运行的结果相加在一起,以获得总体测试开发结果?
有人对此有任何解决方案吗?我已经检查了所有文件名和结果格式,它们似乎是正确的:
文件名: detections_test-dev2017_ALG_results.json
.json内容示例:
[
{
"image_id": 1,
"category_id": 8,
"bbox": [
270.3,
0.9,
367.5,
217.8
],
"score": 0.938004195690155
},
{
"image_id": 1,
"category_id": 3,
"bbox": [
13.3,
122.4,
443.6,
189.3
],
"score": 0.858835756778717
},
{
"image_id": 1,
"category_id": 7,
"bbox": [
13.3,
121.9,
439.6,
191.0
],
"score": 0.6183867454528809
}
]