-
I have to questions. First what e2e accuracy mean? its accuracy for detection and recognition? And det-hmean is only for detection?
And second on wich datasets was this model trained?
-
请问‘_C’在什么地方
-
Hi.
The corresponding link of custom dataset example(https://universityofadelaide.app.box.com/s/fo7odnmqe370btm7sdotqve1c0zsu8p3) cannot access. Could you update it or upload this file?
-
CUDA out of memory. Tried to allocate 136.00 MiB (GPU 3; 10.76 GiB total capacity; 1.72 GiB already allocated; 74.56 MiB free; 2.01 GiB reserved in total by PyTorch)
-
Hi, when I run condinst, I meet the question as follow,
-- Process 1 terminated with the following error:
Traceback (most recent call last):
File "/home/l547/anaconda3/envs/detectron2/lib/python3…
-
Hi:
I follow the instructions of this repo to train ConInst with Base-CondInst config. However, I obtain very poor performance, anything is wrong?
[06/16 16:50:14 d2.data.detection_utils]: TransformGens used in trainin…
-
How to use "Bezier_curve_generation.py" and "generate_bezier_json.py" to generage abcnet train-set?
-
Hi, great work by adelaiDet.
I have pulled and installed the **newest detectron2 & AdelaiDet**.
I can successfully run evaluation on totaltext, and get e2e-hmean 67.1 by using model `tt_attn_R_50`…
-