asheshjain399 / rnnexp Goto Github PK

Hi
I want to ask you about the input and output files (.pik format). How can I access these files correctly.
I would like to see the actual input features style.

Sorry to bother you, but I can not successfully run the 'python generateMotionForecast.py model path'
I don't know why?

I have read the forcastTrajectories.py, and find:

I creat a basedir file saved the h3.6m dir,and checkpoint.pik file is also in there.

I do not know how to run the file.Thank you!

this module, math, is no third-part library. I don`t know why this so. I will appreciate it if you could help me.

>>$ python readData.py 
Traceback (most recent call last):
  File "readData.py", line 1, in <module>
    import numpy as np
  File "/usr/lib/python2.7/dist-packages/numpy/__init__.py", line 180, in <module>
    from . import add_newdocs
  File "/usr/lib/python2.7/dist-packages/numpy/add_newdocs.py", line 13, in <module>
    from numpy.lib import add_newdoc
  File "/usr/lib/python2.7/dist-packages/numpy/lib/__init__.py", line 3, in <module>
    import math
ImportError: No module named math

Hi,

I am trying to download pre-trained SRNN and 3LR models as instructed in there "Pre-trained models of S-RNN, ERD, and LSTM-3LR can be downloaded from here" (readme file) but I can not download the models. Clicking on the links brings me back on the same page and no download starts. It would be great if you could have a look at it.

HI,
Is there a way to display the animation of the generated sequence or training sequence?
I found a script called 'generateAndSaveVideos.m', but I don't know how it works?
Is it possible to save the cdf file to a bvh file by given the skeleton information? If yes, how to do this, and how can I get the skeleton information, because I didn't see any skeleton information from cdf files. Or maybe there is a direct way to display cdf file without need of transferring to bvh format.
Thanks !

I tried to run the simulation using the 'all' maneuver-type, fold_1. I modified the maneuver-rnn.py so it takes in maneuver-type rather than index argument. I'm getting an error during the execution of MultipleRNNsCombined. It says too many values to unpack. Any thoughts on what might be the problem.

python maneuver-rnn.py 'all' 1
(7, 1888, 13)
(7, 1888)
<type 'numpy.float32'>
Number of classes 6
Feature dimension 13
Traceback (most recent call last):
File "maneuver-rnn.py", line 107, in
rnn = MultipleRNNsCombined([layers_1,layers_2],output_layer,softmax_decay_loss,trY,step_size,Adagrad())
File "/home/oolabiyi/brain4cars/NeuralModels/neuralmodels/models/MultipleRNNsCombined.py", line 59, in init
self.train = theano.function([self.X[0],self.X[1],self.Y],self.cost,updates=self.updates)
File "/home/oolabiyi/anaconda/lib/python2.7/site-packages/theano/compile/function.py", line 266, in function
profile=profile)
File "/home/oolabiyi/anaconda/lib/python2.7/site-packages/theano/compile/pfunc.py", line 489, in pfunc
no_default_updates=no_default_updates)
File "/home/oolabiyi/anaconda/lib/python2.7/site-packages/theano/compile/pfunc.py", line 191, in rebuild_collect_shared
for (store_into, update_val) in iter_over_pairs(updates):
ValueError: too many values to unpack

in processdata.py, what does nodeFeatureRanges mean? It seems that the features are just numbers?nodeFeaturesRanges = {} nodeFeaturesRanges['torso'] = range(6) nodeFeaturesRanges['torso'].extend(range(36, 51)) nodeFeaturesRanges['right_arm'] = range(75, 99) nodeFeaturesRanges['left_arm'] = range(51, 75) nodeFeaturesRanges['right_leg'] = range(6, 21) nodeFeaturesRanges['left_leg'] = range(21, 36)

Hi, what a great work you have done here!
Could you show your processing steps from .cdf file(raw file on Human3.6M) to .txt file that could publicly be downloaded on your github page?

Hi!

I'm using the pre-trained models available at https://drive.google.com/open?id=0B7lfjqylzqmMZlI3TUNUUEFQMXc and running generateMotionForecast.py. This produces motion predictions for different activities and models, but I've found that these do not correspond to what is reported in the paper. For reference, here's the figure from the paper that I'm talking about:

But, for example, using the lstm3lr_walking model, the checkpoint.pikforecast_error file contains the following values:

T=0 2.87922000885, 0.053318887949
T=1 3.31045722961, 0.0613047629595
T=2 3.72076749802, 0.0689031034708
T=3 4.20972061157, 0.0779577866197
T=4 4.62205123901, 0.0855935439467
T=5 4.9056763649, 0.0908458605409
T=6 5.15456962585, 0.0954549908638
T=7 5.68943977356, 0.105359993875

T=8 6.14819526672, 0.113855466247
T=9 6.47697734833, 0.119944028556
T=10 6.86927509308, 0.127208799124
T=11 7.25948381424, 0.134434878826
T=12 7.56049823761, 0.140009224415
T=13 7.60584354401, 0.140848949552
T=14 7.81918954849, 0.144799813628
T=15 7.99432945251, 0.148043140769
T=16 8.21197509766, 0.152073606849
T=17 8.22490978241, 0.152313143015
T=18 8.21773910522, 0.152180358768
T=19 8.20940303802, 0.152025982738
T=20 8.21308326721, 0.152094140649
T=21 8.08870410919, 0.14979082346
T=22 7.9909658432, 0.147980853915
T=23 7.93785572052, 0.146997332573
T=24 8.08372688293, 0.149698644876
T=25 8.17058372498, 0.151307106018
T=26 8.29908180237, 0.153686702251
T=27 8.29321861267, 0.15357811749
T=28 8.33865356445, 0.154419511557
T=29 8.29992961884, 0.153702393174
T=30 8.31999206543, 0.154073923826
T=31 8.37398910522, 0.155073866248
T=32 8.47292232513, 0.156905964017
T=33 8.59246826172, 0.159119784832
T=34 8.65988731384, 0.160368278623
T=35 8.66351318359, 0.160435423255
T=36 8.65542507172, 0.160285651684
T=37 8.70272254944, 0.161161527038
T=38 8.90265083313, 0.16486389935
T=39 9.08981990814, 0.168329998851
T=40 9.22410964966, 0.170816838741
T=41 9.25332164764, 0.171357810497
T=42 9.3009595871, 0.172239989042
T=43 9.29813861847, 0.172187745571
T=44 9.26357460022, 0.171547681093
T=45 9.19590568542, 0.170294553041
T=46 9.15723419189, 0.169578418136
T=47 9.24366569519, 0.171178996563
T=48 9.30495262146, 0.172313943505
T=49 9.25953674316, 0.171472907066
T=50 9.24114990234, 0.171132400632
T=51 9.26937294006, 0.171655058861
T=52 9.3104429245, 0.172415614128
T=53 9.19757270813, 0.170325413346
T=54 9.04441356659, 0.167489141226
T=55 8.96823406219, 0.166078403592
T=56 9.00592136383, 0.166776314378
T=57 9.09947776794, 0.168508842587
T=58 9.06608009338, 0.167890369892
T=59 9.1175775528, 0.168844029307
T=60 9.23169708252, 0.170957356691
T=61 9.25059127808, 0.171307250857
T=62 9.23868370056, 0.171086728573
T=63 9.21300506592, 0.170611202717
T=64 9.20988750458, 0.170553475618
T=65 9.30304908752, 0.172278687358
T=66 9.30745029449, 0.17236019671
T=67 9.29339599609, 0.172099933028
T=68 9.21964550018, 0.170734182
T=69 9.22905826569, 0.170908480883
T=70 9.11111068726, 0.168724268675
T=71 9.0918712616, 0.168367981911
T=72 8.92658901215, 0.165307208896
T=73 8.91659736633, 0.165122166276
T=74 8.82111263275, 0.163353934884
T=75 8.90966320038, 0.16499376297
T=76 9.02032756805, 0.167043104768
T=77 9.09782981873, 0.168478325009
T=78 9.22392463684, 0.170813426375
T=79 9.33905029297, 0.172945380211
T=80 9.31301212311, 0.172463193536
T=81 9.44260978699, 0.174863144755
T=82 9.45653438568, 0.17512100935
T=83 9.52670955658, 0.176420554519
T=84 9.64883327484, 0.178682103753
T=85 9.83387374878, 0.182108774781
T=86 9.95151329041, 0.184287279844
T=87 9.91870689392, 0.183679759502
T=88 9.91715335846, 0.18365098536
T=89 10.0150337219, 0.18546359241
T=90 9.95522022247, 0.184355929494
T=91 9.70408630371, 0.179705306888
T=92 9.56737327576, 0.1771735847
T=93 9.58298301697, 0.177462652326
T=94 9.52612495422, 0.176409721375
T=95 9.55842971802, 0.177007958293
T=96 9.53139877319, 0.176507383585
T=97 9.50600910187, 0.176037207246
T=98 9.59951972961, 0.177768886089
T=99 9.80951976776, 0.181657776237

where the left and right columns correspond to skel_err and err_per_dof as computed in forecastTrajectories.py#L124

skel_err = np.mean(np.sqrt(np.sum(np.square((forecasted_motion - trY_forecasting)),axis=2)),axis=1)
err_per_dof = skel_err / trY_forecasting.shape[2]

I find one value to be much worse, and the other to be about 1 order of magnitude better. Do you have any pointers as to what I could be doing wrong?

Hi,

Could you please provide the dataset and checkpoint (or, at least the instruction) for the RNN implemetation of activity-anticipation?

I can't find the dataset folders in the repository or instruction in README file.

path_to_dataset = '/scr/ashesh/activity-anticipation/dataset/{0}'.format(fold)
path_to_checkpoints = '/scr/ashesh/activity-anticipation/checkpoints/{0}'.format(fold)

Thank you so much.

I just download the dataset from https://www.dropbox.com/sh/yndzlk3o90ooq2j/AACWUT8xjabmILM6-rm1_gNAa?dl=0 and then run the maneuver-rnn.py as :

~/RNNexp/anticipatory-rnn/maneuver-anticipation$ python maneuver-rnn.py 960453 fold_1

The error info:
File "maneuver-rnn.py", line 23, in
test_data = cPickle.load(open('{1}/test_data_{0}.pik'.format(index,path_to_dataset)))
IOError: [Errno 2] No such file or directory: '/home/dong/brain4car/brain4cars_data/fold_1/test_data_960453.pik'

how could I get the correct floder path and pik files?
thanks for your help

Hello, great paper and code provided! Thank you! But I have a question about the the input of nodeRNN(torso for instance). The paper says the nodeRNN(torso) concatenates the nodeFeatures(torso) and outputs of corresponding edgeRNNS(torso_input, torso_arm, torso_leg)[AS SEEN IN FIG.4]. But in the implementation(configuration: srnn), I found that there is no nodeFeatures(torso) sent into the nodeRNN(torso). Is that right? Besides, in the code, does the 'torso_input' mean 'torso_torso temporal edgeRNNs'? Very much looking forward to your reply!

Hi, I'm using hyperParameterTuning.py lstm3lr to train lstm3lr model. However, I noticed that the checkpoints doesn't contain the model I just trained.
If I want to get access to the weights of the model I just trained, how can I do that?

Thanks!

Sorry, where can I download the pretrained models for S-RNN, ERD, and LSTM-3LR? The readme has a self-referencing link to https://github.com/asheshjain399/RNNexp/blob/srnn/structural_rnn

Hi Ashesh,
How to run S-RNN on the H3.6m dataset?
Running generateMotionData.py or processdata.py gave me the error:
ImportError: No module named readCRFGraph
Thanks,
Val.

Can you provide the preprocessed data represented with quaternions instead of exponential maps, please?

Thanks

Hi,

Can you please provide the dataset and checkpoint for the RNN implemetation?

I can't find the dataset folders in the repo. I need to correct the following lines in maneuver-rnn.py

path_to_dataset = '/scr/ashesh/brain4cars/dataset/{0}'.format(fold)
path_to_checkpoints = '/scr/ashesh/brain4cars/checkpoints/{0}'.format(fold)

Thanks.

Hi,

I would like to clarify the sources for the dataset and checkpoints for reproducing the maneuver anticipation algorithm. As far as I understand, the dataset can be found here: (https://www.dropbox.com/sh/yndzlk3o90ooq2j/AACWUT8xjabmILM6-rm1_gNAa?dl=0)
But I am not sure about checkpoints. Is this just this path: RNNexp/anticipatory-rnn/maneuver-anticipation/checkpoints/?

Thank you,
Sasha