我从 Github 将 repo 导入到 Colab 当我运行它时没有响应

Question

I import this repo to Colab. The goal of this project is detect vehicles.

import os
os.chdir("/content/drive/uygulama")
!git clone https://github.com/xuannianz/EfficientDet.git
!python3 /content/drive/uygulama/EfficientDet/inference.py

I see a long response. There is no any error. But also there is no any response. When I run it on Anaconda Prompt. I can see video.

2020-04-03 18:16:19.781211: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-04-03 18:16:21.705357: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1
2020-04-03 18:16:21.718914: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.719461: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0 with properties: 
pciBusID: 0000:00:04.0 name: Tesla P100-PCIE-16GB computeCapability: 6.0
coreClock: 1.3285GHz coreCount: 56 deviceMemorySize: 15.90GiB deviceMemoryBandwidth: 681.88GiB/s
2020-04-03 18:16:21.719526: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-04-03 18:16:21.719593: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10
2020-04-03 18:16:21.719625: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10
2020-04-03 18:16:21.719657: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10
2020-04-03 18:16:21.719687: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10
2020-04-03 18:16:21.720621: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10
2020-04-03 18:16:21.720704: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-04-03 18:16:21.720808: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.721346: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.721859: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1703] Adding visible gpu devices: 0
2020-04-03 18:16:21.727125: I tensorflow/core/platform/profile_utils/cpu_utils.cc:102] CPU Frequency: 2200000000 Hz
2020-04-03 18:16:21.727305: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1d65100 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-04-03 18:16:21.727352: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version
2020-04-03 18:16:21.816983: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.817861: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1d652c0 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-04-03 18:16:21.817893: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Tesla P100-PCIE-16GB, Compute Capability 6.0
2020-04-03 18:16:21.818109: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.818693: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0 with properties: 
pciBusID: 0000:00:04.0 name: Tesla P100-PCIE-16GB computeCapability: 6.0
coreClock: 1.3285GHz coreCount: 56 deviceMemorySize: 15.90GiB deviceMemoryBandwidth: 681.88GiB/s
2020-04-03 18:16:21.818752: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-04-03 18:16:21.818776: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10
2020-04-03 18:16:21.818804: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10
2020-04-03 18:16:21.818827: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10
2020-04-03 18:16:21.818846: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10
2020-04-03 18:16:21.818891: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10
2020-04-03 18:16:21.818909: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-04-03 18:16:21.819005: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.819632: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:21.820138: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1703] Adding visible gpu devices: 0
2020-04-03 18:16:21.820201: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-04-03 18:16:22.362078: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-04-03 18:16:22.362140: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1108]      0 
2020-04-03 18:16:22.362153: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1121] 0:   N 
2020-04-03 18:16:22.362389: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:22.363088: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-04-03 18:16:22.363704: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:39] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.
2020-04-03 18:16:22.363748: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1247] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14974 MB memory) -> physical GPU (device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0)

Then, nothing happens. I expect to see a video frame by frame. Why there is no any response?

This is inference.py

import cv2
import json
import numpy as np
import os
import time
import glob

from model import efficientdet
from utils import preprocess_image, postprocess_boxes
from utils.draw_boxes import draw_boxes


def main():
    os.environ['CUDA_VISIBLE_DEVICES'] = '0' #ekran kartı olsaydı 0 -> 1 olacaktı.

    phi = 0 # çözünürlük indisi. phi ne ile eğittiyseniz ona göre ayarlayacaksınız.
    weighted_bifpn = True
    model_path = '/content/drive/uygulama/EfficientDet/efficientdet-d0.h5' #modelin pathi ve kendisi. 
    #çözünürlük olarak 512 yi mi seçtik indisi 0 yaparak: evet.
    image_sizes = (512, 640, 768, 896, 1024, 1280, 1408)
    image_size = image_sizes[phi]
    # coco classes
    classes = {value['id'] - 1: value['name'] for value in json.load(open('/content/drive/uygulama/EfficientDet/coco_90.json', 'r')).values()} # bu datasetler 90 farklı nesneye göre 
    #oluşturulmuş datasetler siz eğer sadece araba, kamyon gibi şeyleri almak istiyorsanız. coco_90.json bu dosyaya gidip aradığınız nesnenin labelına bakacaksınız.
    # örneğin car 3 bisiklet 2 bunu bir if else yapısı ile aşağıda ayrıştıracaksınız.
    num_classes = 90
    score_threshold = 0.3 # modelin verdiği score'un eşiklenmesi. Örneğin bir nesneye 0.2 dediyse onu almayacak.
    #anladım bunun bir formülü var mı deneme yanılma mı? : Modelin başarısına göre siz ayarlayacaksınız. Deneme yanılma biraz.
    colors = [np.random.randint(0, 256, 3).tolist() for _ in range(num_classes)]
    _, model = efficientdet(phi=phi,
                            weighted_bifpn=weighted_bifpn, #false yapıp dene
                            num_classes=num_classes,
                            score_threshold=score_threshold)
    model.load_weights(model_path, by_name=True)

    cap = cv2.VideoCapture('t.mp4')

    while cap.isOpened():
        ret,image = cap.read()
        src_image = image.copy()
        # BGR -> RGB
        image = image[:, :, ::-1]
        h, w = image.shape[:2]

        image, scale = preprocess_image(image, image_size=image_size)
        # run network
        start = time.time()
        boxes, scores, labels = model.predict_on_batch([np.expand_dims(image, axis=0)]) # modelin çalıştığı yer
        boxes, scores, labels = np.squeeze(boxes), np.squeeze(scores), np.squeeze(labels)
        print(time.time() - start)
        boxes = postprocess_boxes(boxes=boxes, scale=scale, height=h, width=w)

        # select indices which have a score above the threshold
        indices = np.where(scores[:] > score_threshold)[0] # eşiklemenin yapıldığı yer.

        # select those detections
        boxes = boxes[indices]
        labels = labels[indices]
        #buradaki labels hangi cisme ait olduğunu gösterir. Bu değer 3 ise o arabadır. Umarım anlatabilmişimdir.
        #şöyle mi olacak? if label == 3 ise alt kısmı gerçkleştir?: Aynen öyle. Tabi bu sizin kodlama becerinize bağlı olarak uğraştırabilir. :)
        # 0.93 yazan confidence değeri.
        # kodlama becerisi derken? basit bir if else yapısı gibi duruyor. o yüzden sordum? : Tabi basit. Ama indisleri iyi seçmek lazım. Gözünüz korkmasın.

        if labels[0]==2 or labels[0]==3:
            draw_boxes(src_image, boxes, scores, labels, colors, classes)

            cv2.namedWindow('image', cv2.WINDOW_NORMAL)
            cv2.imshow('image', src_image)
            cv2.waitKey(0)
    cap.release()
    cv2.destroyAllWindows()  

if __name__ == '__main__':
    main()

Answer 1

The problem is here

cv2.namedWindow('image', cv2.WINDOW_NORMAL)
cv2.imshow('image', src_image)
cv2.waitKey(0)

It tries to open a new window, then wait for your keyboard. But Colab is run in the cloud. It cannot access your window or keyboard. You can modify the code to save the image and display it in another way.