v0.10.0

[fcakyon]

New Features

- Layer.ai integration

from sahi import AutoDetectionModel

detection_model = AutoDetectionModel.from_layer("layer/yolov5/models/yolov5s")

result = get_sliced_prediction(
    "image.jpeg",
    detection_model,
    slice_height = 512,
    slice_width = 512,
    overlap_height_ratio = 0.2,
    overlap_width_ratio = 0.2
)

- Huggingface Transformers object detectors

from sahi.model import HuggingfaceDetectionModel

detection_model = HuggingfaceDetectionModel(
    model_path="facebook/detr-resnet-50",
    image_size=640,
    confidence_threshold=0.5
)

result = get_sliced_prediction(
    "image.jpeg",
    detection_model,
    slice_height = 512,
    slice_width = 512,
    overlap_height_ratio = 0.2,
    overlap_width_ratio = 0.2
)

- TorchVision object detectors

import torchvision
from sahi.model import TorchVisionDetectionModel

model = torchvision.models.detection.fasterrcnn_resnet50_fpn(pretrained=True)

detection_model = TorchVisionDetectionModel(
    model=model,
    image_size=640,
    confidence_threshold=0.5
)

result = get_sliced_prediction(
    "image.jpeg",
    detection_model,
    slice_height = 512,
    slice_width = 512,
    overlap_height_ratio = 0.2,
    overlap_width_ratio = 0.2
)

- Support for exporting predictions in COCO format

from sahi.utils.coco import Coco, CocoImage, CocoAnnotation, CocoPrediction
from sahi.utils.file import save_json
from pycocotools.cocoeval import COCOeval
from pycocotools.coco import COCO

coco_obj = Coco()

# add n images to coco_obj
for _ in range(n):
    image = CocoImage(**kwargs)
    
    # add n annotations to the image
    for _ in ange(n):
        image.add_annotation(CocoAnnotation(**kwargs))
    
    # add n predictions to the image
    for _ in range(n)
        image.add_prediction(CocoPrediction(**kwargs))
    
    # add image to coco object
    coco_obj.add_image(image)

# export ground truth annotations
coco_gt = coco_obj.json
save_json(coco_gt , "ground_truth.json")

# export predictions 
coco_predictions = coco_obj.prediction_array
save_json(coco_predictions, "predictions.json")

coco_ground_truth = COCO(annotation_file="coco_dataset.json")
coco_predictions = coco_ground_truth.loadRes("coco_predictions.json")
coco_evaluator = COCOeval(coco_ground_truth, coco_predictions, "bbox")
coco_evaluator.evaluate()
coco_evaluator.accumulate()
coco_evaluator.summarize()

What's Changed

• refactor torch utils by @fcakyon in refactor torch utils #468
• add automodel structure for unified model loading by @fcakyon in add automodel structure for unified model loading #469
• add support to instantiate a DetectionModel from layer by @mecevit in add support to instantiate a DetectionModel from layer #462
• refactor automodel by @fcakyon in refactor automodel #470
• update layer versions in workflows by @fcakyon in update layer versions in workflows #471
• update version to 0.10.0 by @fcakyon in update version to 0.10.0 #474
• Added support for exporting predictions in COCO format by @PushpakBhoge in Added support for exporting predictions in COCO format #465
• update contributors in readme by @fcakyon in update contributors in readme #477
• update device priority for Detectron2DetectionModel by @fcakyon in update device priority for Detectron2DetectionModel #479
• fix pickle export for video by @fcakyon in fix pickle export for video #481
• update continuous integration by @fcakyon in update continuous integration #483
• refactor import and torch utils by @fcakyon in refactor import and torch utils #484
• make detectionmodel classes more explicit in automodel by @fcakyon in make detectionmodel classes more explicit in automodel #485
• utilize check_requirements in several modules by @fcakyon in utilize check_requirements in several modules #487
• update package versions in workflows by @fcakyon in update package versions in workflows #488
• add support for huggingface transformers object detectors by @devrimcavusoglu in add support for huggingface transformers object detectors #475
• add torchvision detector support by @fcakyon in add torchvision detector support #486
• remove legacy image_size parameter by @kadirnar in remove legacy image_size parameter #494
• AutoDetectionModel can be imported from sahi by @fcakyon in AutoDetectionModel can be imported from sahi #498
• add python3.6 support by @fcakyon in add python3.6 support #489
• refactor exception handling by @kadirnar in refactor exception handling #499
• improve requirements and import handling by @fcakyon in improve requirements and import handling #502

New Contributors

• @mecevit made their first contribution in add support to instantiate a DetectionModel from layer #462
• @PushpakBhoge made their first contribution in Added support for exporting predictions in COCO format #465

Full Changelog: 0.9.4...0.10.0

---

This discussion was created from the release v0.10.0.