sleepat-11
sleepat-11
i changed the input size to 224×224,and compare RepLKNet with resnet101, it seems that resnet101 is larger than RepLKNet-13, left is RepLKNet-13, right is resnet101 
for example ,i want to use the quaternion to present arbitrary orientation from [distance_sensor](https://mavlink.io/en/messages/common.html#DISTANCE_SENSOR). My question is should i Generate a Custom MAVLink Dialect follow this [link](https://mavlink.io/en/getting_started/generate_libraries.html) or there are...
作者你好,感谢你的工作,我使用在sceneflow上预训练的模型并在kitti上finetune,请问在用kitti训练网络的时候,验证性能没有划分验证集吗?我看evaluate_stereo.py代码中仍然是在训练集上评估性能,没有划分验证集,您是如何挑选最好的模型的呢?目前我看kitti leadboard不再支持调试提交结果了