-
Hi teams,
I am using 2021.3 lgsvl but encounter that any Apollo cars with ground truth sensor cannot detect the pedestrian (NPC detection works fine).
![Screenshot from 2023-05-04 11-20-38](https…
-
This is a great job.
Hello, when reproducing your work, I followed the steps you provided and the demo file ran smoothly. However, I encountered a problem during training.
This is my running log:
+…
-
I git the code, and train it with 4 GPUs, the final validation result is below, and the result match the author's log:
Car AP@0.70, 0.70, 0.70:
bbox AP:96.1959, 90.2071, 89.5205
bev AP:90.3205, 88…
-
Firstly thank you for the great and intuitive work for collision detection. I am developing an application where I need to find the collision point between the moving robot and detected/tracked pedest…
Rak-r updated
6 months ago
-
## I install used the source code
I installed successfully before I execute git pull commond
after I execute the git pull, error occured
```sh
$ sudo python setup.py install
[sudo] password for w…
-
DENSE_HEAD:
NAME: CenterHead
CLASS_AGNOSTIC: False
CLASS_NAMES_EACH_HEAD: [
['Vehicle', 'Pedestrian', 'Cyclist']
]
SHARED_CONV_CHANNEL…
-
@ALL
Hey, anyone who knows why the evaluation results (i.e. Car_easy =0.00....) like as follows:
Evaluating detections...
Running: ./src/dataset/kitti-eval/cpp/evaluate_object ./data/KITTI/traini…
-
I just pull the darkflow today and run it using both python and Annocoda, Mincoda.
I also followed this tutorial
https://keponk.wordpress.com/2017/12/07/siraj-darkflow/
I have spent a whole …
-
@lArkl So you acquired .weights and .cfg files after training, right?
If your .cfg and .weights files are particularly fine-tuned for pedestrian only, could you upload them for us to use right away…
-
Retinanet uses resnet50 as backbone to train Kitti data sets. How much is the result of Moderate, Easy and Hard on car, pedestrian and cyclist? Has anyone trained?
This is my training result. The tes…