# Copyright 2022 AI Singapore
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""🔲 One-stage Object Detection model."""
from typing import Any, Dict, List, Optional, Union
import cv2
import numpy as np
from peekingduck.pipeline.nodes.abstract_node import AbstractNode
from peekingduck.pipeline.nodes.model.yolov4 import yolo_model
[docs]class Node(AbstractNode):
"""Initializes and uses YOLO model to infer bboxes from image frame.
The yolo node is capable of detecting objects from 80 categories. It uses
YOLOv4-tiny by default and can be changed to using YOLOv4. The table of
categories can be found :ref:`here <general-object-detection-ids>`.
Inputs:
|img_data|
Outputs:
|bboxes_data|
|bbox_labels_data|
|bbox_scores_data|
Configs:
model_type (:obj:`str`): **{"v4", "v4tiny"}, default="v4tiny"**. |br|
Defines the type of YOLO model to be used.
weights_parent_dir (:obj:`Optional[str]`): **default = null**. |br|
Change the parent directory where weights will be stored by
replacing ``null`` with an absolute path to the desired directory.
num_classes (:obj:`int`): **default = 80**. |br|
Maximum number of objects to be detected.
detect (:obj:`List[Union[int, str]]`): **default = [0]**. |br|
List of object class names or IDs to be detected. To detect all classes,
refer to the :ref:`tech note <general-object-detection-ids>`.
max_output_size_per_class (:obj:`int`): **default = 50**. |br|
Maximum number of detected instances for each class in an image.
max_total_size (:obj:`int`): **default = 50**. |br|
Maximum total number of detected instances in an image.
iou_threshold (:obj:`float`): **[0, 1], default = 0.5**. |br|
Overlapping bounding boxes above the specified IoU (Intersection
over Union) threshold are discarded.
score_threshold (:obj:`float`): **[0, 1], default = 0.2**. |br|
Bounding box with confidence score less than the specified
confidence score threshold is discarded.
References:
YOLOv4: Optimal Speed and Accuracy of Object Detection:
https://arxiv.org/pdf/2004.10934v1.pdf
Model weights trained by
https://github.com/hunglc007/tensorflow-yolov4-tflite
Inference code adapted from https://github.com/zzh8829/yolov3-tf2
.. versionchanged:: 1.2.0 |br|
``yolo_iou_threshold`` is renamed to ``iou_threshold``. |br|
``yolo_score_threshold`` is renamed to ``score_threshold``.
"""
def __init__(self, config: Dict[str, Any] = None, **kwargs: Any) -> None:
super().__init__(config, node_path=__name__, **kwargs)
self.model = yolo_model.YOLOModel(self.config)
def run(self, inputs: Dict[str, Any]) -> Dict[str, Any]:
"""Reads the image input and returns the bboxes of the specified
objects chosen to be detected.
Args:
inputs (dict): Dictionary of inputs with key "img".
Returns:
outputs (dict): bbox output in dictionary format with keys
"bboxes", "bbox_labels", and "bbox_scores".
"""
image = cv2.cvtColor(inputs["img"], cv2.COLOR_BGR2RGB)
bboxes, labels, scores = self.model.predict(image)
bboxes = np.clip(bboxes, 0, 1)
outputs = {"bboxes": bboxes, "bbox_labels": labels, "bbox_scores": scores}
return outputs
def _get_config_types(self) -> Dict[str, Any]:
return {
"detect": List[Union[int, str]],
"iou_threshold": float,
"max_output_size_per_class": int,
"max_total_size": int,
"model_type": str,
"num_classes": int,
"score_threshold": float,
"weights_parent_dir": Optional[str],
}