soldierofhell

velocity PID in carla_ackermann_control

1

Hi, Could someone please confirm that my understanding of comment attached to velocity PID control in ackermann control is right. Here's the comment: > Run the PID control loop for...

VIO

2

Hi @clydemcqueen, Tello IMU data aren't standard raw IMU measurments. Is it possible to do VIO just using monocular video + Tello IMU messages? Regards,

[Feature Request] (partially) known dynamics

2

Hi, Usually in practice we have some dynamic model which at least approximate the true environment x_{t+1} = f(x_t, u_t) (e.g from Newton laws). My question is how to include...

Hi, it seems like this flag is reversed in this if https://github.com/yongxinw/GSDT/blob/0aad69429e4c4692b6b82ce7d0c087bfcbb3d4eb/src/track_gnn.py#L74 (roi_align is in else:) Regards,

detectron2 Panoptic-FPN

4

From the README: ``` One additional repo will be introduced in October 2020: mseg-panoptic: provides Panoptic-FPN and Mask-RCNN training, based on Detectron2 ``` I wonder if this is still planned?...

Hi @homangab, Thanks for the effort to boost the performance of CEM optimization in the right (gradient) direction. However reading the code I don't see (probably should be somwhere [here](https://github.com/homangab/gradcem/blob/02a8b36269704ab7e4c1207b6420cc788286fd67/mpc/gradcem.py#L62))...

why decoding starts from 3rd position?

3

Hi, I wonder what is the theoretical basis for starting decoding from 3rd position. I'm referring to this line: `ctc_decode = bknd.ctc_decode(y_pred[:, 2:, :], input_length=np.ones(shape[0])*shape[1])[0][0]` In image_ocr.py example on keras...

Motion prediction: protos vs. tf.Examples

6

Hi, In the description of the dataset on https://waymo.com/open/data/motion/ it states: > Each tf.Example proto contains the same information as the Scenario protos described above, but all data has been...

K.set_learning_phase(0)

2

Hi, I don't get it why in training.py there's a line 7, which forces the learning_phase to 0 (=test)? `K.set_learning_phase(0)` As far as I know for example BatchNormalization() layer behaves...

CNN - other architectures and transfer learning

2

Has anyone tried to replace default simple CNN structure with pretrained architectures like VGG, InceptionV3, etc.? On one hand the typical text image (like plates) is different in nature than...