This is a mashup of IceVision's "Custom Parser" example and their "Getting Started (Object Detection)" notebooks, to analyze SPNet Real dataset, for which I generated bounding boxes. -- shawley, July 1, 2021
Installing IceVision and IceData
If on Colab run the following cell, else check the installation instructions
#try:
# !wget https://raw.githubusercontent.com/airctic/icevision/master/install_colab.sh
# !chmod +x install_colab.sh && ./install_colab.sh
#except:
# print("Ignore the error messages and just keep going")
import torch, re
tv, cv = torch.__version__, torch.version.cuda
tv = re.sub('\+cu.*','',tv)
TORCH_VERSION = 'torch'+tv[0:-1]+'0'
CUDA_VERSION = 'cu'+cv.replace('.','')
print(f"TORCH_VERSION={TORCH_VERSION}; CUDA_VERSION={CUDA_VERSION}")
print(f"CUDA available = {torch.cuda.is_available()}, Device count = {torch.cuda.device_count()}, Current device = {torch.cuda.current_device()}")
print(f"Device name = {torch.cuda.get_device_name()}")
#!pip install -qq mmcv-full=="1.3.8" -f https://download.openmmlab.com/mmcv/dist/{CUDA_VERSION}/{TORCH_VERSION}/index.html --upgrade
#!pip install mmdet -qq
from icevision.all import *
import pandas as pd
We're going to be using a small sample of the chess dataset, the full dataset is offered by roboflow here
#!rm -rf /root/.icevision/data/espiownage-cyclegan
#data_url = "https://anonymized.machine.com/~drscotthawley/spnet_sample-master.zip"
#data_dir = icedata.load_data(data_url, 'spnet_sample') / 'spnet_sample-master'
# can use public espiownage cyclegan dataset:
#data_url = 'https://anonymized.machine.com/~drscotthawley/espiownage-cyclegan.tgz'
#data_dir = icedata.load_data(data_url, 'espiownage-cyclegan') / 'espiownage-cyclegan'
# or local data already there:
from pathlib import Path
data_dir = Path('/home/drscotthawley/datasets/espiownage-cyclegan')
In this task we were given a .csv file with annotations, let's take a look at that.
!!! danger "Important"
Replace source with your own path for the dataset directory.
df = pd.read_csv(data_dir / "bboxes/annotations.csv")
df.head()
At first glance, we can make the following assumptions:
- Multiple rows with the same filename, width, height
- A label for each row
- A bbox [xmin, ymin, xmax, ymax] for each row
Once we know what our data provides we can create our custom Parser.
set(np.array(df['label']).flatten())
#df['label'] = ["Object"]*len(df)# "_"+df['label'].apply(str) # force label to be string-like
df['label'] /= 2
#df.head()
df['label'] = df['label'].apply(int)
print(set(np.array(df['label']).flatten()))
df['label'] = "_"+df['label'].apply(str)+"_"
df.head()
df['label'] = 'AN' # antinode
df.head()
The first step is to create a template record for our specific type of dataset, in this case we're doing standard object detection:
template_record = ObjectDetectionRecord()
Now use the method generate_template that will print out all the necessary steps we have to implement.
Parser.generate_template(template_record)
We can copy the template and use it as our starting point. Let's go over each of the methods we have to define:
__init__: What happens here is completely up to you, normally we have to pass some reference to our data,data_dirin our case.__iter__: This tells our parser how to iterate over our data, each item returned here will be passed toparse_fieldsaso. In our case we calldf.itertuplesto iterate over alldfrows.__len__: How many items will be iterating over.imageid: Should return aHashable(int,str, etc). In our case we want all the dataset items that have the samefilenameto be unified in the same record.parse_fields: Here is where the attributes of the record are collected, the template will suggest what methods we need to call on the record and what parameters it expects. The parameteroit receives is the item returned by__iter__.
!!! danger "Important"
Be sure to pass the correct type on all record methods!
class BBoxParser(Parser):
def __init__(self, template_record, data_dir):
super().__init__(template_record=template_record)
self.data_dir = data_dir
self.df = pd.read_csv(data_dir / "bboxes/annotations.csv")
#self.df['label'] /= 2
#self.df['label'] = self.df['label'].apply(int)
#self.df['label'] = "_"+self.df['label'].apply(str)+"_"
self.df['label'] = 'AN' # make them all the same object
self.class_map = ClassMap(list(self.df['label'].unique()))
def __iter__(self) -> Any:
for o in self.df.itertuples():
yield o
def __len__(self) -> int:
return len(self.df)
def record_id(self, o) -> Hashable:
return o.filename
def parse_fields(self, o, record, is_new):
if is_new:
record.set_filepath(self.data_dir / 'images' / o.filename)
record.set_img_size(ImgSize(width=o.width, height=o.height))
record.detection.set_class_map(self.class_map)
record.detection.add_bboxes([BBox.from_xyxy(o.xmin, o.ymin, o.xmax, o.ymax)])
record.detection.add_labels([o.label])
Let's randomly split the data and parser with Parser.parse:
parser = BBoxParser(template_record, data_dir)
train_records, valid_records = parser.parse()
Let's take a look at one record:
show_record(train_records[5], display_label=False, figsize=(14, 10))
train_records[0]
# size is set to 384 because EfficientDet requires its inputs to be divisible by 128
image_size = 384
train_tfms = tfms.A.Adapter([*tfms.A.aug_tfms(size=image_size, presize=512), tfms.A.Normalize()])
valid_tfms = tfms.A.Adapter([*tfms.A.resize_and_pad(image_size), tfms.A.Normalize()])
# Datasets
train_ds = Dataset(train_records, train_tfms)
valid_ds = Dataset(valid_records, valid_tfms)
samples = [train_ds[0] for _ in range(3)]
show_samples(samples, ncols=3)
model_type = models.mmdet.retinanet
backbone = model_type.backbones.resnet50_fpn_1x(pretrained=True)
selection = 1
extra_args = {}
if selection == 0:
model_type = models.mmdet.retinanet
backbone = model_type.backbones.resnet50_fpn_1x
elif selection == 1:
# The Retinanet model is also implemented in the torchvision library
model_type = models.torchvision.retinanet
backbone = model_type.backbones.resnet50_fpn
elif selection == 2:
model_type = models.ross.efficientdet
backbone = model_type.backbones.tf_lite0
# The efficientdet model requires an img_size parameter
extra_args['img_size'] = image_size
elif selection == 3:
model_type = models.ultralytics.yolov5
backbone = model_type.backbones.small
# The yolov5 model requires an img_size parameter
extra_args['img_size'] = image_size
model_type, backbone, extra_args
model = model_type.model(backbone=backbone(pretrained=True), num_classes=len(parser.class_map), **extra_args)
train_dl = model_type.train_dl(train_ds, batch_size=8, num_workers=4, shuffle=True)
valid_dl = model_type.valid_dl(valid_ds, batch_size=8, num_workers=4, shuffle=False)
model_type.show_batch(first(valid_dl), ncols=4)
metrics = [COCOMetric(metric_type=COCOMetricType.bbox)]
learn = model_type.fastai.learner(dls=[train_dl, valid_dl], model=model, metrics=metrics)
learn.lr_find(end_lr=0.005)
# For Sparse-RCNN, use lower `end_lr`
# learn.lr_find(end_lr=0.005)
learn.fine_tune(60, 1e-4, freeze_epochs=2)
model_type.show_results(model, valid_ds, detection_threshold=.5)
preds = model_type.predict(model, valid_ds, keep_images=True)
show_preds(preds=preds[0:10])
