I'm following the tutorial on using your own template images to do object 3D pose tracking, but I'm trying to get it working on Ubuntu 20.04 with a live webcam stream.
I was able to successfully make my index .pb file with extracted KNIFT features from my custom images.
It seems the next thing to do is load the provided template matching graph (in mediapipe/graphs/template_matching/template_matching_desktop.pbtxt) (replacing the index_proto_filename of the BoxDetectorCalculator with my own index file), and run it on a video input stream to track my custom object.
I was hoping that would be easiest to do in python, but am running into dependency problems.
(I installed mediapipe python with pip3 install mediapipe)
First, I couldn't find how to directly load a .pbtxt file as a graph in the mediapipe python API, but that's ok. I just load the text it contains and use that.
template_matching_graph_filepath=os.path.abspath("~/mediapipe/mediapipe/graphs/template_matching/template_matching_desktop.pbtxt")
graph = mp.CalculatorGraph(graph_config=open(template_matching_graph_filepath).read())
But I get missing calculator targets.
No registered object with name: OpenCvVideoDecoderCalculator; Unable to find Calculator "OpenCvVideoDecoderCalculator"
or
[libprotobuf ERROR external/com_google_protobuf/src/google/protobuf/text_format.cc:309] Error parsing text-format mediapipe.CalculatorGraphConfig: 54:70: Could not find type "type.googleapis.com/mediapipe.TfLiteInferenceCalculatorOptions" stored in google.protobuf.Any.
It seems similar to this troubleshooting case but, since I'm not trying to compile an application, I'm not sure how to link in the missing calculators.
How to I make the mediapipe python API aware of these graphs?
UPDATE:
I made decent progress by adding the graphs that the template_matching depends on to the cc_library deps of the mediapipe/python/BUILD file
cc_library(
name = "builtin_calculators",
deps = [
"//mediapipe/calculators/image:feature_detector_calculator",
"//mediapipe/calculators/image:image_properties_calculator",
"//mediapipe/calculators/video:opencv_video_decoder_calculator",
"//mediapipe/calculators/video:opencv_video_encoder_calculator",
"//mediapipe/calculators/video:box_detector_calculator",
"//mediapipe/calculators/tflite:tflite_inference_calculator",
"//mediapipe/calculators/tflite:tflite_tensors_to_floats_calculator",
"//mediapipe/calculators/util:timed_box_list_id_to_label_calculator",
"//mediapipe/calculators/util:timed_box_list_to_render_data_calculator",
"//mediapipe/calculators/util:landmarks_to_render_data_calculator",
"//mediapipe/calculators/util:annotation_overlay_calculator",
...
I also modified solution_base.py so it knows about BoxDetector's options.
from mediapipe.calculators.video import box_detector_calculator_pb2
...
CALCULATOR_TO_OPTIONS = {
'BoxDetectorCalculator':
box_detector_calculator_pb2
.BoxDetectorCalculatorOptions,
Then I rebuilt and installed mediapipe python from source with:
~/mediapipe$ python3 setup.py install --link-opencv
Then I was able to make my own class derived from SolutionBase
from mediapipe.python.solution_base import SolutionBase
class ObjectTracker(SolutionBase):
"""Process a video stream and output a video with edges of templates highlighted."""
def __init__(self,
object_knift_index_file_path):
super().__init__(binary_graph_path=object_pose_estimation_binary_file_path,
calculator_params={"BoxDetector.index_proto_filename": object_knift_index_file_path},
)
def process(self, image: np.ndarray) -> NamedTuple:
return super().process(input_data={'input_video':image})
ot = ObjectTracker(object_knift_index_file_path="/path/to/my/object_knift_index.pb")
Finally, I process a video frame from a cv2.VideoCapture
cv_video = cv2.VideoCapture(0)
result, frame = cv_video.read()
input_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
res = ot.process(image=input_frame)
So close! But I run into this error which I just don't know what to do with.
/usr/local/lib/python3.8/dist-packages/mediapipe/python/solution_base.py in process(self, input_data)
326 if data.shape[2] != RGB_CHANNELS:
327 raise ValueError('Input image must contain three channel rgb data.')
--> 328 self._graph.add_packet_to_input_stream(
329 stream=stream_name,
330 packet=self._make_packet(input_stream_type,
RuntimeError: Graph has errors:
Calculator::Open() for node "BoxDetector" failed: ; Error while reading file: /usr/local/lib/python3.8/dist-packages/
Looks like CalculatorNode::OpenNode() is trying to open the python API install path as a file. Maybe it has to do with the default_context. I have no idea where to go from here. :(
Related
I have a tensorflow "graph-model" consisting of a model.json and several .bin files. In javascript I am able to read those files using
const weights = browser.runtime.getURL("web_model/model.json");
tf.loadGraphModel(weights)
However I would like to be able to use this model in python, in order to process the results better.
When I try to load the model in python with
new_model = keras.models.load_model('./web_model/model.json')
I get the following error:
File "h5py/h5f.pyx", line 106, in h5py.h5f.open
OSError: Unable to open file (file signature not found)
I don't understand, since the javascript code is able to run the model, I think python should be able to do the same as well. What am I doing wrong ?
I need to run a custom GluonCV object detection module on Android.
I already fine-tuned the model (ssd_512_mobilenet1.0_custom) on a custom dataset, I tried running inference with it (loading the .params file produced during the training) and everything works perfectly on my computer. Now, I need to export this to Android.
I was referring to this answer to figure out the procedure, there are 3 suggested options:
You can use ONNX to convert models to other runtimes, for example [...] NNAPI for Android
You can use TVM
You can use SageMaker Neo + DLR runtime [...]
Regarding the first one, I converted my model to ONNX.
However, in order to use it with NNAPI, it is necessary to convert it to daq. In the repository, they provide a precomplied AppImage of onnx2daq to make the conversion, but the script returns an error. I checked the issues section, and they report that "It actually fails for all onnx object detection models".
Then, I gave a try to DLR, since it's suggested to be the easiest way.
As I understand, in order to use my custom model with DLR, I would first need to compile it with TVM (which also covers the second point mentioned in the linked post). In the repo, they provide a Docker image with some conversion scripts for different frameworks.
I modified the 'compile_gluoncv.py' script, and now I have:
#!/usr/bin/env python3
from tvm import relay
import mxnet as mx
from mxnet.gluon.model_zoo.vision import get_model
from tvm_compiler_utils import tvm_compile
shape_dict = {'data': (1, 3, 300, 300)}
dtype='float32'
ctx = [mx.cpu(0)]
classes_custom = ["CML_mug"]
block = get_model('ssd_512_mobilenet1.0_custom', classes=classes_custom, pretrained_base=False, ctx=ctx)
block.load_parameters("ep_035.params", ctx=ctx) ### this is the file produced by training on the custom dataset
for arch in ["arm64-v8a", "armeabi-v7a", "x86_64", "x86"]:
sym, params = relay.frontend.from_mxnet(block, shape=shape_dict, dtype=dtype)
func = sym["main"]
func = relay.Function(func.params, relay.nn.softmax(func.body), None, func.type_params, func.attrs)
tvm_compile(func, params, arch, dlr_model_name)
However, when I run the script it returns the error:
ValueError: Model ssd_512_mobilenet1.0_custom is not supported. Available options are
alexnet
densenet121
densenet161
densenet169
densenet201
inceptionv3
mobilenet0.25
mobilenet0.5
mobilenet0.75
mobilenet1.0
mobilenetv2_0.25
mobilenetv2_0.5
mobilenetv2_0.75
mobilenetv2_1.0
resnet101_v1
resnet101_v2
resnet152_v1
resnet152_v2
resnet18_v1
resnet18_v2
resnet34_v1
resnet34_v2
resnet50_v1
resnet50_v2
squeezenet1.0
squeezenet1.1
vgg11
vgg11_bn
vgg13
vgg13_bn
vgg16
vgg16_bn
vgg19
vgg19_bn
Am I doing something wrong? Is this thing even possible?
As a side note, after this I'd need to deploy on Android a pose detection model (simple_pose_resnet18_v1b) and an activity recognition one (i3d_nl10_resnet101_v1_kinetics400) as well.
You actually can run GluonCV model directly on Android with Deep Java Library (DJL)
What you need to do is:
hyridize your GluonCV model and save as MXNet model
Build MXNet engine for android, MXNET already support Android build
Include MXNet shared library into your android project
Use DJL in your android project, you can follow this DJL Android demo for PyTorch
The error message is self-explanatory - there is no model "ssd_512_mobilenet1.0_custom" supported by mxnet.gluon.model_zoo.vision.get_model. You are confusing GluonCV's get_model with MXNet Gluon's get_model.
Replace
block = get_model('ssd_512_mobilenet1.0_custom',
classes=classes_custom, pretrained_base=False, ctx=ctx)
with
import gluoncv
block = gluoncv.model_zoo.get_model('ssd_512_mobilenet1.0_custom',
classes=classes_custom, pretrained_base=False, ctx=ctx)
I am trying to work with the quite recently published tensorflow_dataset API to train a Keras model on the Open Images Dataset. The dataset is about 570 GB in size. I downloaded the data with the following code:
import tensorflow_datasets as tfds
import tensorflow as tf
open_images_dataset = tfds.image.OpenImagesV4()
open_images_dataset.download_and_prepare(download_dir="/notebooks/dataset/")
After the download was complete, the connection to my jupyter notebook somehow interrupted but the extraction seemed to be finished as well, at least all downloaded files had a counterpart in the "extracted" folder. However, I am not able to access the downloaded data now:
tfds.load(name="open_images_v4", data_dir="/notebooks/open_images_dataset/extracted/", download=False)
This only gives the following error:
AssertionError: Dataset open_images_v4: could not find data in /notebooks/open_images_dataset/extracted/. Please make sure to call dataset_builder.download_and_prepare(), or pass download=True to tfds.load() before trying to access the tf.data.Dataset object.
When I call the function download_and_prepare() it only downloads the whole dataset again.
Am I missing something here?
Edit:
After the download the folder under "extracted" has 18 .tar.gz files.
This is with tensorflow-datasets 1.0.1 and tensorflow 2.0.
The folder hierarchy should be like this:
/notebooks/open_images_dataset/extracted/open_images_v4/0.1.0
All the datasets have a version. Then the data could be loaded like this.
ds = tf.load('open_images_v4', data_dir='/notebooks/open_images_dataset/extracted', download=False)
I didn't have open_images_v4 data. I put cifar10 data into a folder named open_images_v4 to check what folder structure tensorflow_datasets was expecting.
The solution to this was to also use the "data_dir" parameter when initializing the dataset:
builder = tfds.image.OpenImagesV4(data_dir="/raid/openimages/dataset")
builder.download_and_prepare(download_dir="/raid/openimages/dataset")
This way the dataset is donwloaded and extracted in the same directory. Before, it was (for me unnoticeably) extracting to the default directory, which is under /home/.../. That's what caused the error, as there wasn't enough space left under my home directory.
After the extraction, the folder structure is exactly as Manoj-Mohan described.
Above solution haven't worked for me.
builder = tfds.builder(name='folder_name', data_dir=data_dir)
builder.download_and_prepare(download_dir="/home/...")
ds = builder.as_dataset()
I am currently trying to get a trained TF seq2seq model working with Tensorflow.js. I need to get the json files for this. My input is a few sentences and the output is "embeddings". This model is working when I read in the checkpoint however I can't get it converted for tf.js. Part of the process for conversion is to get my latest checkpoint frozen as a protobuf (pb) file and then convert that to the json formats expected by tensorflow.js.
The above is my understanding and being that I haven't done this before, it may be wrong so please feel free to correct if I'm wrong in what I have deduced from reading.
When I try to convert to the tensorflow.js format I use the following command:
sudo tensorflowjs_converter --input_format=tf_frozen_model
--output_node_names='embeddings'
--saved_model_tags=serve
./saved_model/model.pb /web_model
This then displays the error listed in this post:
ValueError: Input 0 of node Variable/Assign was passed int32 from
Variable:0 incompatible with expected int32_ref.
One of the problems I'm running into is that I'm really not even sure how to troubleshoot this. So I was hoping that perhaps one of you maybe had some guidance or maybe you know what my issue may be.
I have upped the code I used to convert the checkpoint file to protobuf at the link below. I then added to the bottom of the notebook an import of that file that is then providing the same error I get when trying to convert to tensorflowjs format. (Just scroll to the bottom of the notebook)
https://github.com/xtr33me/textsumToTfjs/blob/master/convert_ckpt_to_pb.ipynb
Any help would be greatly appreciated!
Still unsure as to why I was getting the above error, however in the end I was able to resolve this issue by just switching over to using TF's SavedModel via tf.saved_model. A rough example of what worked for me can be found below should anyone in the future run into something similar. After saving out the below model, I was then able to perform the tensorflowjs_convert call on it and export the correct files.
if first_iter == True: #first time through
first_iter = False
#Lets try saving this badboy
cwd = os.getcwd()
path = os.path.join(cwd, 'simple')
shutil.rmtree(path, ignore_errors=True)
inputs_dict = {
"batch_decoder_input": tf.convert_to_tensor(batch_decoder_input)
}
outputs_dict = {
"batch_decoder_output": tf.convert_to_tensor(batch_decoder_output)
}
tf.saved_model.simple_save(
sess, path, inputs_dict, outputs_dict
)
print('Model Saved')
#End save model code
After following this tutorial on summaries and TensorBoard, I've been able to successfully save and look at data with TensorBoard. Is it possible to open this data with something other than TensorBoard?
By the way, my application is to do off-policy learning. I'm currently saving each state-action-reward tuple using SummaryWriter. I know I could manually store/train on this data, but I thought it'd be nice to use TensorFlow's built in logging features to store/load this data.
As of March 2017, the EventAccumulator tool has been moved from Tensorflow core to the Tensorboard Backend. You can still use it to extract data from Tensorboard log files as follows:
from tensorboard.backend.event_processing.event_accumulator import EventAccumulator
event_acc = EventAccumulator('/path/to/summary/folder')
event_acc.Reload()
# Show all tags in the log file
print(event_acc.Tags())
# E. g. get wall clock, number of steps and value for a scalar 'Accuracy'
w_times, step_nums, vals = zip(*event_acc.Scalars('Accuracy'))
Easy, the data can actually be exported to a .csv file within TensorBoard under the Events tab, which can e.g. be loaded in a Pandas dataframe in Python. Make sure you check the Data download links box.
For a more automated approach, check out the TensorBoard readme:
If you'd like to export data to visualize elsewhere (e.g. iPython
Notebook), that's possible too. You can directly depend on the
underlying classes that TensorBoard uses for loading data:
python/summary/event_accumulator.py (for loading data from a single
run) or python/summary/event_multiplexer.py (for loading data from
multiple runs, and keeping it organized). These classes load groups of
event files, discard data that was "orphaned" by TensorFlow crashes,
and organize the data by tag.
As another option, there is a script
(tensorboard/scripts/serialize_tensorboard.py) which will load a
logdir just like TensorBoard does, but write all of the data out to
disk as json instead of starting a server. This script is setup to
make "fake TensorBoard backends" for testing, so it is a bit rough
around the edges.
I think the data are encoded protobufs RecordReader format. To get serialized strings out of files you can use py_record_reader or build a graph with TFRecordReader op, and to deserialize those strings to protobuf use Event schema. If you get a working example, please update this q, since we seem to be missing documentation on this.
I did something along these lines for a previous project. As mentioned by others, the main ingredient is tensorflows event accumulator
from tensorflow.python.summary import event_accumulator as ea
acc = ea.EventAccumulator("folder/containing/summaries/")
acc.Reload()
# Print tags of contained entities, use these names to retrieve entities as below
print(acc.Tags())
# E. g. get all values and steps of a scalar called 'l2_loss'
xy_l2_loss = [(s.step, s.value) for s in acc.Scalars('l2_loss')]
# Retrieve images, e. g. first labeled as 'generator'
img = acc.Images('generator/image/0')
with open('img_{}.png'.format(img.step), 'wb') as f:
f.write(img.encoded_image_string)
You can also use the tf.train.summaryiterator: To extract events in a ./logs-Folder where only classic scalars lr, acc, loss, val_acc and val_loss are present you can use this GIST: tensorboard_to_csv.py
Chris Cundy's answer works well when you have less than 10000 data points in your tfevent file. However, when you have a large file with over 10000 data points, Tensorboard will automatically sampling them and only gives you at most 10000 points. It is a quite annoying underlying behavior as it is not well-documented. See https://github.com/tensorflow/tensorboard/blob/master/tensorboard/backend/event_processing/event_accumulator.py#L186.
To get around it and get all data points, a bit hacky way is to:
from tensorboard.backend.event_processing.event_accumulator import EventAccumulator
class FalseDict(object):
def __getitem__(self,key):
return 0
def __contains__(self, key):
return True
event_acc = EventAccumulator('path/to/your/tfevents',size_guidance=FalseDict())
It looks like for tb version >=2.3 you can streamline the process of converting your tb events to a pandas dataframe using tensorboard.data.experimental.ExperimentFromDev().
It requires you to upload your logs to TensorBoard.dev, though, which is public. There are plans to expand the capability to locally stored logs in the future.
https://www.tensorflow.org/tensorboard/dataframe_api
You can also use the EventFileLoader to iterate through a tensorboard file
from tensorboard.backend.event_processing.event_file_loader import EventFileLoader
for event in EventFileLoader('path/to/events.out.tfevents.xxx').Load():
print(event)
Surprisingly, the python package tb_parse has not been mentioned yet.
From documentation:
Installation:
pip install tensorflow # or tensorflow-cpu pip install -U tbparse # requires Python >= 3.7
Note: If you don't want to install TensorFlow, see Installing without TensorFlow.
We suggest using an additional virtual environment for parsing and plotting the tensorboard events. So no worries if your training code uses Python 3.6 or older versions.
Reading one or more event files with tbparse only requires 5 lines of code:
from tbparse import SummaryReader
log_dir = "<PATH_TO_EVENT_FILE_OR_DIRECTORY>"
reader = SummaryReader(log_dir)
df = reader.scalars
print(df)