Commit Graph

60 Commits

Author SHA1 Message Date
lorenzo
f30fd24b2b add exception for Nonetype 2019-07-01 16:26:53 +02:00
lorenzo
b0c3f51897 fix matches in case of no ground truth 2019-07-01 16:20:00 +02:00
Lorenzo Bertoni
019b6b0fad
make utils in torch and remove redundant functions (#3)
* Add precision metrics

* add mode gt_all and change default threshold

* add cyclists

* add iou matrix'
'

* add cyclists only in training phase

* add dropout in model name

* small typos

* small typo

* fix error on uv_boxes

* change default mode from gt_ped to gt

* 2 decimals

* fix name bug

* refactor prepare_pif_kps

* corrected get_keypoints_batch

* add pixel to camera for 3d vectors

* preprocessing in torch

* return original outputs

* Skeleton for post_process

* baseline version for post processing

* add keypoints torch in post_processing

* cleaning misc

* add reorder_matches

* update preprocess with get_iou_matches

* fix indices

* remove aa

* temp

* skeleton kitti_generate

* skeleton kitti_generate (2)

* refactor file

* remove old get_input_data

* refactor geometric eval

* refactor geometric eval(2)

* temp

* refactor geometric

* change saving order for txts

* update pixel to camera

* update depth

* Fix pixel to camera

* add xyz_from_distance

* use new function

* fix std_ale calculation in eval

* remove debug points
2019-06-28 18:33:58 +02:00
Lorenzo Bertoni
2aea30cb7d
add matrix IoU evaluation (#2)
* add boxes_3d and fix bug on kk

* delete stereo script

* add get_iou_matrix

* add statistics

* Add pytest

* add iou_matrix

* refactor evaluation kitti

* add box size constraint on moderate

* refactor evaluation

* pylint check
2019-06-24 08:45:39 +02:00
Lorenzo Bertoni
2b9177ea06
Update README.md
Add link paper
2019-06-20 18:44:16 +02:00
Lorenzo Bertoni
468c23c6f1
Add link paper 2019-06-20 18:27:18 +02:00
Lorenzo Bertoni
a61886ad5f
Add link paper 2019-06-20 18:26:07 +02:00
lorenzo
52f5592889 update factory_for_gt 2019-06-20 18:21:48 +02:00
Lorenzo Bertoni
86438189a7
refactor predict (#1)
* add stereo arg (#1)

* stereo skeleton

* skeleton stereo (#2)

* skeleton stereo (#3)

* refactor for stereo

* stereo running

* modify stereo parameters

* remove warnings as errors

* add p2_right

* add factory file

* add p2_right

* add scaling factor to intrinsic matrix

* Refactor predict

* refactor predict 2

* temp

* add uppercase constants

* add uppercase constants

* working predict

* add todo

* Add person_sitting as option to uncomment

* turn off stereo

* remove stereo

* pylint corrections

* add break point

* temp

* fix small bug

* Add compatibility for pifpaf and monoloco networks

* pylint fix
2019-06-20 15:43:38 +02:00
lorenzo
d941a5bdc7 add video link 2019-06-14 09:58:37 +02:00
lorenzo
bf7dcd6a3c fix argument 2019-05-28 08:52:53 +02:00
lorenzo
086a4d909a refactor class 2019-05-28 08:52:10 +02:00
lorenzo
ad252e76e7 specify stereo method 2019-05-24 18:22:19 +02:00
lorenzo
3771555f36 add assertion for arg --dir_ann 2019-05-24 18:22:01 +02:00
lorenzo
e7a9807518 add license 2019-05-24 17:28:57 +02:00
lorenzo
cf9b118345 add model links 2019-05-24 17:18:19 +02:00
lorenzo
b7b9cf2912 Update readme (1) 2019-05-24 17:05:07 +02:00
lorenzo
f142645dd3 add generic image predictions 2019-05-24 17:00:17 +02:00
lorenzo
1f57f8f235 add description for generic images 2019-05-24 16:37:35 +02:00
lorenzo
0a8f41f987 add 2 aspect ratios for kitti and nuscenes 2019-05-24 16:20:22 +02:00
lorenzo
dfdddae43a remove y_scale parameter 2019-05-24 16:11:38 +02:00
lorenzo
9c547f418b add combined for different dataset 2019-05-24 14:22:09 +02:00
lorenzo
afc439862d update model name 2019-05-24 13:18:14 +02:00
lorenzo
1a979d3d4d include epistemic uncertainty in vis limits 2019-05-24 13:15:47 +02:00
lorenzo
0a56b439d1 pylint refactor 2019-05-22 19:35:24 +02:00
lorenzo
eeea8945fb pylint refactor (1) 2019-05-21 18:34:57 +02:00
lorenzo
02ac6626d6 refactor kitti eval 2019-05-21 15:13:34 +02:00
lorenzo
e9df8c9c1e remove useless arguments 2019-05-21 15:10:53 +02:00
lorenzo
3c7d07d081 add print method 2019-05-21 15:03:46 +02:00
lorenzo
6f3379e394 add factory method 2019-05-21 14:45:23 +02:00
lorenzo
496e147c2a refactor figures 2019-05-21 14:25:35 +02:00
lorenzo
762163877b change dic_stats format 2019-05-21 12:12:57 +02:00
lorenzo
45bb9df863 add class variables 2019-05-21 11:44:50 +02:00
lorenzo
632e138cd1 add function for statistics 2019-05-21 11:41:47 +02:00
lorenzo
3a42536cef add internal methods 2019-05-21 11:23:45 +02:00
lorenzo
521a04ece5 refactor methods 2019-05-21 11:20:35 +02:00
lorenzo
67bc780677 refactor __init__(2) 2019-05-21 11:11:07 +02:00
lorenzo
6fbc496702 refactor __init (1) 2019-05-21 10:42:33 +02:00
lorenzo
738095bc8c add function for split_training 2019-05-21 10:30:21 +02:00
lorenzo
3c6c305606 remove class imports 2019-05-21 10:26:59 +02:00
lorenzo
755688818f add class variables 2019-05-21 10:21:39 +02:00
lorenzo
e96b37f8c8 add factory method 2019-05-21 09:57:05 +02:00
lorenzo
d756678518 fix pylint(2) 2019-05-16 12:26:42 +02:00
lorenzo
3243da72cd fix pylint 2019-05-16 10:30:52 +02:00
lorenzo
bf8fbf5234 small fix 2019-05-13 18:53:00 +02:00
lorenzo
2222d12e20 remove no-calib argument 2019-05-13 17:59:13 +02:00
lorenzo
3cbc818059 add required fields (2) 2019-05-13 17:44:16 +02:00
lorenzo
456bbc0620 add required fields 2019-05-13 17:41:10 +02:00
lorenzo
511b9ae650 remove unused figures (2) 2019-05-13 16:06:32 +02:00
lorenzo
da4a351e95 remove unused figures 2019-05-13 16:06:19 +02:00