ina-foss / inafaceanalyzer Goto Github PK

may require new functions from image

pytorch & tensorflow together ???
https://github.com/DavidDoukhan/face_recognition

As part of the JOSS review:

As outlined in the reviewer instructions, the documentation (in this case: the readme) should include a "A statement of need". Perhaps you could include a short paragraph (copied or paraphrased from the manuscript)?

Add tests with the following image
https://pixabay.com/photos/retro-beauty-hair-bob-bobbed-2118473/

ability perform eye detection and rotation
save the 224*244 rotated &cropped image => ability to extract the correct features

https://scanner.run/
https://github.com/DavidDoukhan/pytorch-DEX

This may be useful for knowing if video had the appropriate length

As part of the JOSS review:

As far as I can see, no community guidelines on how to contribute/report issues/seek help are listed. A short paragraph on how to do this (through github issues/PRs, I assume) would suffice, I think.

WARNING:tensorflow:5 out of the last 9 calls to <function Model.make_predict_function..predict_function at 0x7f94e0427d30> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.

Hi, I'm reviewing the package for JOSS (see here) and I'll create separate issues for the things I encounter during the review process.

Considering the author contributions: @DavidDoukhan clearly did most of the work, but I can't find any contributions in terms of code/documentation of Thomas Petit. In contrast, as is mentioned under 'CREDITS' in the readme, @ZohraRezgui was responsible for the prototype of inaFaceAnalyzer, but is not listed as a co-author. Could the decision to include Thomas Petit but exclude Zohra Rezgui be explained?

I'm trying to install inaFaceAnalyzer on my macOS (Python 3.7). I have tried Installing from pypi and building from source but getting error messgae in both:
ERROR: Could not find a version that satisfies the requirement onnxruntime-gpu (from inaFaceAnalyzer==0.6.2+6.g9c345a5) (from versions: none)
ERROR: No matching distribution found for onnxruntime-gpu (from inaFaceAnalyzer==0.6.2+6.g9c345a5).

I installed onnxruntime (only present for cpu version for mac) from pypi but still no luck.
Is there a way to run the library on mac without the GPU support. Thanks!

As part of the JOSS review:

Although I really appreciate the useful tutorial notebooks, I think they do not replace a complete API documentation of the Python interface. Adding a documentation of at least the most important package functionality (e.g., the VideoAnalyzer, VideoKeyframes, VideoTracking, and ImageAnalyzer classes) would be great!

Make sure you ran the unit tests before submitting a bug report and tell us if and where they fail.
You can use the program test_inaFaceAnalyzer.py at the root of this repository.

Describe the bug

A clear and concise description of what the bug is.

To Reproduce

• ina_face_analyzer.py --engine image -i -o results.csv
• Note that the image I used to test is a PDF file of a Matplotlib plot. I can't share the exact image but I hope the description can help you generate a test image for your use

Expected behavior

File should be processed without throwing an exception

Current behavior

Tell us what happens instead of the expected behavior.
Include any log/backtrace here.
Traceback (most recent call last):
File "$HOME/miniconda3/envs/face/bin/ina_face_analyzer.py", line 139, in
df = engine(args.input)
File "$HOME/miniconda3/envs/face/lib/python3.9/site-packages/inaFaceAnalyzer/inaFaceAnalyzer.py", line 217, in call
return self._process_stream(stream, self.face_detector)
File "$HOME/miniconda3/envs/face/lib/python3.9/site-packages/inaFaceAnalyzer/inaFaceAnalyzer.py", line 161, in _process_stream
for iframe, frame in stream_iterator:
File "$HOME/miniconda3/envs/face/lib/python3.9/site-packages/inaFaceAnalyzer/opencv_utils.py", line 84, in image_iterator
yield f, imread_rgb(f, verbose)
File "$HOME/miniconda3/envs/face/lib/python3.9/site-packages/inaFaceAnalyzer/opencv_utils.py", line 128, in imread_rgb
frame = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
cv2.error: OpenCV(4.6.0) /io/opencv/modules/imgproc/src/color.cpp:182: error: (-215:Assertion failed) !_src.empty() in function 'cvtColor'

Environment

Please provide the following details on your environment :

• OS and distribution: Linux Ubuntu 18.04

• Python version : 3.9.12

• Tensorflow version : 2.9.1

• output of command pip freeze :
  absl-py==1.1.0
  astunparse==1.6.3
  av==9.2.0
  cachetools==5.2.0
  certifi==2022.5.18.1
  charset-normalizer==2.0.12
  Cheetah3==3.2.6.post1
  cycler==0.11.0
  dlib==19.24.0
  flatbuffers==1.12
  fonttools==4.33.3
  gast==0.4.0
  google-auth==2.7.0
  google-auth-oauthlib==0.4.6
  google-pasta==0.2.0
  grpcio==1.46.3
  h5py==3.7.0
  idna==3.3
  importlib-metadata==4.11.4
  inaFaceAnalyzer==0.6.1
  joblib==1.1.0
  keras==2.9.0
  Keras-Preprocessing==1.1.2
  kiwisolver==1.4.2
  libclang==14.0.1
  Markdown==3.3.7
  matplotlib==3.5.2
  numpy==1.22.4
  oauthlib==3.2.0
  onnxruntime-gpu==1.11.1
  opencv-contrib-python==4.6.0.66
  opt-einsum==3.3.0
  packaging==21.3
  pandas==1.4.2
  Pillow==9.1.1
  protobuf==3.19.4
  pyasn1==0.4.8
  pyasn1-modules==0.2.8
  pyparsing==3.0.9
  Pyro4==4.82
  python-dateutil==2.8.2
  pytz==2022.1
  requests==2.28.0
  requests-oauthlib==1.3.1
  rsa==4.8
  scikit-learn==1.1.1
  scipy==1.8.1
  serpent==1.40
  six==1.16.0
  sklearn==0.0
  tensorboard==2.9.1
  tensorboard-data-server==0.6.1
  tensorboard-plugin-wit==1.8.1
  tensorflow==2.9.1
  tensorflow-estimator==2.9.0
  tensorflow-io-gcs-filesystem==0.26.0
  termcolor==1.1.0
  threadpoolctl==3.1.0
  typing_extensions==4.2.0
  urllib3==1.26.9
  Werkzeug==2.1.2
  wrapt==1.14.1
  zipp==3.8.0

• Amount of available RAM :

• Running on GPU or CPU :

• output of command nvidia-smi (if GPU is used) :

• CUDA and CuDNN versions (if GPU is used) :

• installation type (from sources, from pypi, from dockerhub ?) :

• inaFaceAnalyzer version :

Additional context

I am following your README as a reviewer for your OSS contribution. See openjournals/joss-reviews#4210

set min dim to 300 to avoid distortion during detection

https://www.pyimagesearch.com/2018/10/29/multi-object-tracking-with-dlib/