Add README, clean requirements and notebook

README.md

@@ -0,0 +1,62 @@
+# ADLC
+
+As of right now, this just includes a script for generating a dataset from images and annotations and a notebook for preliminary testing on using a CNN for object detection.
+
+Here is a sample output of `visualize_detections`
+
+![four img with bounding box](./img/output.png)
+
+Since the images are really high resolution, the lines cover the targets, but the boxes are accurate.
+
+## Data Annotation
+
+For now, I have done data annotation using OpenCVs GUI tool [`opencv_annotation`](https://docs.opencv.org/4.x/dc/d88/tutorial_traincascade.html#Preparation-of-the-training-data), which uses a XYWH bounding-box format. There are annotations in `data/annotation_238.txt`.
+
+To include the corresponding images, you will need to download them from the Kraken computer and place them in `data/flight_238/*.jpg`. They are located in `/RAID/Flights/Flight_238/*.jpg`.
+
+## Setup Development Environment
+
+### Using a Conda/Mamba Environment
+
+Create a conda environment using:
+
+```sh
+conda env create --file ncsuadlc_condaenv.yaml -n ncsuadlc
+conda activate ncsuadlc
+
+# Some requirements are only up-to-date on PyPi
+pip install -r ncsuadlc_pipreqs.txt
+```
+
+### Pip Only
+
+```sh
+pip install -r requirements.txt
+```
+
+## Using CuDNN Acceleration on VLC
+
+NCSU provides VLCs with RTX 2080 GPUs that can be used for training the CNN quickly. CUDA is already installed on these systems but you will need to install CuDNN as well:
+
+```sh
+sudo apt-get install libcudnn8=8.8.0.121-1+cuda12.1
+sudo apt-get install libcudnn8-dev=8.8.0.121-1+cuda12.1
+sudo apt-get install libcudnn8-samples=8.8.0.121-1+cuda12.1
+```
+
+To check that CuDNN was set up correctly, run built-in test suite:
+
+```sh
+cp -r /usr/src/cudnn_samples_v8/ $HOME
+cd $HOME/cudnn_samples_v8/mnistCUDNN
+make clean && make
+sudo apt-get install libfreeimage3 libfreeimage-dev
+make clean && make
+./mnistCUDNN
+```
+
+You will also need to make sure that Tensorflow has needed GPU dependencies using:
+
+```sh
+pip install tensorflow[and-cuda]
+```

adlc_util.py

@@ -3,26 +3,26 @@
 
 CLASS_MAPPING = {0:"target"}
 
-def visualize_dataset(inputs, value_range, rows, cols, bounding_box_format, offset=1):
- """
- Unused, but works
- """
- it = iter(inputs.take(offset))
+def visualize_detections(model, dataset, bounding_box_format, offset=1):
+ it = iter(dataset.take(offset))
 
  for _ in range(offset):
- inputs = next(it)
+ images, y_true = next(it)
 
- images, bounding_boxes = inputs["images"], inputs["bounding_boxes"]
+ # images, y_true = next(iter(dataset.take(1)))
+ y_pred = model.predict(images)
+ y_pred = bounding_box.to_ragged(y_pred)
  visualization.plot_bounding_box_gallery(
  images,
- value_range=value_range,
- rows=rows,
- cols=cols,
- y_true=bounding_boxes,
- scale=10,
- line_thickness=1,
- font_scale=0.7,
+ value_range=(0, 255),
  bounding_box_format=bounding_box_format,
+ y_true=y_true,
+ y_pred=y_pred,
+ scale=20,
+ rows=2,
+ cols=2,
+ show=True,
+ font_scale=0.7,
  class_mapping=CLASS_MAPPING,
  )
 

ncsuadlc_condaenv.yaml

@@ -0,0 +1,48 @@
+name: ncsuadlc
+channels:
+- conda-forge
+dependencies:
+- _libgcc_mutex==0.1
+- _openmp_mutex==4.5
+- bzip2==1.0.8
+- ca-certificates
+- click
+- empy
+- ipykernel
+- lark
+- ld_impl_linux-64==2.40
+- libblas==3.9.0
+- libcblas==3.9.0
+- libexpat==2.5.0
+- libffi==3.4.2
+- libgcc-ng==13.2.0
+- libgfortran-ng==13.2.0
+- libgfortran5==13.2.0
+- libgomp==13.2.0
+- liblapack==3.9.0
+- libnsl==2.0.0
+- libopenblas==0.3.24
+- libsqlite==3.43.0
+- libstdcxx-ng==13.2.0
+- libuuid==2.38.1
+- libzlib==1.2.13
+- ncurses==6.4
+- numpy==1.25.2
+- openssl
+- pandas
+- pillow
+- pip==23.2.1
+- protobuf
+- pycocotools
+- python==3.11.5
+- python_abi==3.11
+- readline==8.2
+- scikit-learn
+- setuptools==68.1.2
+- tk==8.6.12
+- tqdm
+- transforms3d==0.4.1
+- tzdata==2023c
+- wheel==0.41.2
+- xz==5.2.6
+

ncsuadlc_pipreqs.txt

@@ -0,0 +1,116 @@
+absl-py==1.4.0
+array-record==0.5.0
+astor==0.8.1
+asttokens==2.4.0
+astunparse==1.6.3
+atomicwrites==1.4.1
+backcall==0.2.0
+backports.functools-lru-cache==1.6.5
+cachetools==5.3.2
+certifi==2023.7.22
+charset-normalizer==3.3.1
+click==8.1.7
+colorama==0.4.6
+comm==0.1.4
+contourpy==1.1.1
+cycler==0.12.1
+Cython==3.0.5
+debugpy==1.8.0
+decorator==5.1.1
+dm-tree==0.1.8
+empy==3.3.4
+etils==1.5.2
+exceptiongroup==1.1.3
+executing==1.2.0
+flatbuffers==23.5.26
+fonttools==4.43.1
+fsspec==2023.10.0
+gast==0.5.4
+google-auth==2.23.4
+google-auth-oauthlib==1.0.0
+google-pasta==0.2.0
+googleapis-common-protos==1.61.0
+grpcio==1.59.2
+h5py==3.10.0
+idna==3.4
+importlib-metadata==6.8.0
+importlib-resources==6.1.0
+ipykernel==6.25.2
+ipython==8.16.1
+jedi==0.19.1
+joblib==1.3.2
+Js2Py==0.74
+jupyter_client==8.4.0
+jupyter_core==5.4.0
+keras==2.14.0
+keras-core==0.1.7
+keras-cv==0.6.4
+kiwisolver==1.4.5
+lark==1.1.7
+libclang==16.0.6
+Markdown==3.5.1
+markdown-it-py==3.0.0
+MarkupSafe==2.1.3
+matplotlib==3.8.0
+matplotlib-inline==0.1.6
+mdurl==0.1.2
+ml-dtypes==0.2.0
+munkres==1.1.4
+namex==0.0.7
+nest-asyncio==1.5.8
+oauthlib==3.2.2
+opencv-python==4.8.1.78
+opt-einsum==3.3.0
+packaging==23.2
+pandas==2.1.2
+parso==0.8.3
+pexpect==4.8.0
+pickleshare==0.7.5
+Pillow==10.1.0
+platformdirs==3.5.1
+promise==2.3
+prompt-toolkit==3.0.39
+protobuf==3.20.3
+psutil==5.9.5
+ptyprocess==0.7.0
+pure-eval==0.2.2
+pyasn1==0.5.0
+pyasn1-modules==0.3.0
+pycocotools==2.0.6
+Pygments==2.16.1
+pyjsparser==2.7.1
+pyparsing==3.1.1
+python-dateutil==2.8.2
+pytz==2023.3.post1
+pyzmq==25.1.1
+regex==2023.10.3
+requests==2.31.0
+requests-oauthlib==1.3.1
+rich==13.6.0
+rsa==4.9
+scikit-learn==1.3.2
+scipy==1.11.3
+six==1.16.0
+stack-data==0.6.2
+tensorboard==2.14.1
+tensorboard-data-server==0.7.2
+tensorflow==2.14.0
+tensorflow-datasets==4.9.3
+tensorflow-estimator==2.14.0
+tensorflow-io-gcs-filesystem==0.34.0
+tensorflow-metadata==1.14.0
+termcolor==2.3.0
+tfds-nightly==4.9.3.dev202310060044
+threadpoolctl==3.2.0
+toml==0.10.2
+tornado==6.3.3
+tqdm==4.66.1
+traitlets==5.11.2
+typing_extensions==4.8.0
+tzdata==2023.3
+tzlocal==5.1
+urllib3==2.0.7
+wcwidth==0.2.8
+Werkzeug==3.0.1
+wrapt==1.14.1
+zipp==3.17.0