test_singleExpert.py

from detectron2.engine import DefaultTrainer ,DefaultPredictor
import os,cv2
# import some common detectron2 utilities
import torch
from detectron2.utils.visualizer import Visualizer ,ColorMode
from detectron2 import model_zoo
from detectron2.config import get_cfg
from dataloader.dataset import Dataset
from detectron2.evaluation import COCOEvaluator, inference_on_dataset
from detectron2.checkpoint import DetectionCheckpointer
from detectron2.data import build_detection_test_loader
from detectron2.modeling import build_model
if __name__=="__main__":
    args="DepthJetQhd/" # change to ImagesQhd/
    x = Dataset(args,set='val')
    cfg = get_cfg()
    model = "COCO-Detection/faster_rcnn_R_50_C4_3x.yaml"
    cfg.merge_from_file(model_zoo.get_config_file(model))
    cfg.DATASETS.TRAIN = ("InOutDoorDepth_val",)
    print("InOutDoorDepth_val")
    cfg.DATASETS.TEST = ()
    cfg.OUTPUT_DIR = 'output'
    cfg.DATALOADER.NUM_WORKERS = 2
    cfg.MODEL.WEIGHTS = 'output/model_0024999.pth'
    cfg.SOLVER.IMS_PER_BATCH = 1
    cfg.SOLVER.BASE_LR = 0.00025  # pick a good LR
    cfg.SOLVER.MAX_ITER = 10000    # 300 iterations seems good enough for this toy dataset; you will need to train longer for a practical dataset
    cfg.MODEL.ROI_HEADS.BATCH_SIZE_PER_IMAGE = 50   # faster, and good enough for this toy dataset (default: 512)
    cfg.MODEL.ROI_HEADS.NUM_CLASSES = 1  # only has one class (ballon). (see https://detectron2.readthedocs.io/tutorials/datasets.html#update-the-config-for-new-datasets)
    # NOTE: this config means the number of classes, but a few popular unofficial tutorials incorrect uses num_classes+1 here.
    # x.visualize()
    # os.makedirs(cfg.OUTPUT_DIR, exist_ok=True)
    # trainer = DefaultTrainer(cfg)
    # trainer.resume_or_load(resume=False)
    # trainer.train()
    pred = DefaultPredictor(cfg)
    # checkpointer2 = DetectionCheckpointer(model)
    # checkpointer2.load(cfg.MODEL.WEIGHTS)
    # model.eval()
    with torch.no_grad():
        # outputs = gn(rgb)[0]
        im = cv2.imread('docs/imagehd.png')

        image1 = cv2.resize(im, (1920, 1080))
        outputs = pred(image1)
        v = Visualizer(image1,
                       scale=0.5,
                       instance_mode=ColorMode.IMAGE_BW
                       # remove the colors of unsegmented pixels. This option is only available for segmentation models
                       )
        print(outputs)
        out = v.draw_instance_predictions(outputs["instances"].to("cpu"))
        cv2.imshow('test', out.get_image()[:, :, ::-1])
        cv2.waitKey(0)
    # evaluator = COCOEvaluator("InOutDoorDepth_val", ("bbox"), False, output_dir="output/")
    # val_loader = build_detection_test_loader(cfg, "InOutDoorDepth_val")
    # print(inference_on_dataset(model, val_loader, evaluator))