'list' object has no attribute 'masks'

[codinglearningnovice]

Search before asking

• I have searched the YOLOv8 issues and discussions and found no similar questions.

Question

i am working in colab, i am trying to access a video, in all frame of video, i want to segment the person in the frame, convert his image to black, then every other image in the background to white. the snippet of my code is this

while(count < TRAIN_SIZE):
try:
ret, frame = cap.read()

if currentFrame % FRAME_SKIP == 0:
    count += 1
    if count % int(TRAIN_SIZE/10) == 0:
        print(str((count/TRAIN_SIZE)*100)+"% done")


    # Perform human segmentation
    results = model(frame)

    person_masks = results.masks[results.boxes.cls == 0]
    person_mask_3ch = cv2.cvtColor(person_masks, cv2.COLOR_GRAY2BGR)
            masked_frame = cv2.bitwise_and(frame, person_mask_3ch)

    inverted_mask = cv2.bitwise_not(person_mask_3ch)

    result_frame = cv2.bitwise_and(masked_frame, inverted_mask)

    resized_frame = cv2.resize(result_frame, (output_width, output_height))

    name = 'trydata/resized_frame.jpg' + str(count) + '.jpg'

    cv2.imwrite(name, resized_frame)

    video.write(resized_frame.astype('uint8'))

except Exception as e:
print(e)
break
currentFrame += 1
print(str(count)+" Frames collected")
cap.release()

video.release()

but i keep geting these errors

WARNING ⚠️ 'source' is missing. Using 'source=/usr/local/lib/python3.10/dist-packages/ultralytics/assets'.

image 1/2 /usr/local/lib/python3.10/dist-packages/ultralytics/assets/bus.jpg: 640x480 4 persons, 1 bus, 1 skateboard, 12.0ms
image 2/2 /usr/local/lib/python3.10/dist-packages/ultralytics/assets/zidane.jpg: 384x640 2 persons, 1 tie, 8.8ms
Speed: 2.9ms preprocess, 10.4ms inference, 2.4ms postprocess per image at shape (1, 3, 384, 640)
'list' object has no attribute 'masks'

Additional

No response

[github-actions]

👋 Hello @codinglearningnovice, thank you for your interest in Ultralytics YOLOv8 🚀! We recommend a visit to the Docs for new users where you can find many Python and CLI usage examples and where many of the most common questions may already be answered.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Join the vibrant Ultralytics Discord 🎧 community for real-time conversations and collaborations. This platform offers a perfect space to inquire, showcase your work, and connect with fellow Ultralytics users.

Install

Pip install the ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8.

pip install ultralytics

Environments

YOLOv8 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all Ultralytics CI tests are currently passing. CI tests verify correct operation of all YOLOv8 Modes and Tasks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

@codinglearningnovice hello,

Thank you for reaching out and providing detailed information about your issue. It looks like you're encountering an error because the results object is a list of Results objects, and you're trying to access the masks attribute directly from the list.

To resolve this, you need to iterate over the results list and then access the masks attribute from each Results object. Here's a modified version of your code snippet that should work:

while(count < TRAIN_SIZE):
    try:
        ret, frame = cap.read()

        if currentFrame % FRAME_SKIP == 0:
            count += 1
            if count % int(TRAIN_SIZE/10) == 0:
                print(str((count/TRAIN_SIZE)*100) + "% done")

            # Perform human segmentation
            results = model(frame)

            for result in results:
                person_masks = result.masks[result.boxes.cls == 0]
                person_mask_3ch = cv2.cvtColor(person_masks, cv2.COLOR_GRAY2BGR)
                masked_frame = cv2.bitwise_and(frame, person_mask_3ch)

                inverted_mask = cv2.bitwise_not(person_mask_3ch)

                result_frame = cv2.bitwise_and(masked_frame, inverted_mask)

                resized_frame = cv2.resize(result_frame, (output_width, output_height))

                name = 'trydata/resized_frame.jpg' + str(count) + '.jpg'

                cv2.imwrite(name, resized_frame)

                video.write(resized_frame.astype('uint8'))

    except Exception as e:
        print(e)
        break

    currentFrame += 1

print(str(count) + " Frames collected")
cap.release()
video.release()

Additionally, please ensure that you are using the latest versions of torch and ultralytics. You can upgrade them using the following commands:

pip install --upgrade torch ultralytics

If the issue persists, please provide a minimum reproducible example so we can investigate further. You can find more details on how to create one here.

I hope this helps! If you have any further questions, feel free to ask. 😊

[codinglearningnovice]

thanks for your reply, tried this, it doesnt give me the result, it issues this error below

0: 384x640 1 person, 190.0ms
Speed: 3.1ms preprocess, 190.0ms inference, 3.8ms postprocess per image at shape (1, 3, 384, 640)
OpenCV(4.8.0) 👎 error: (-5:Bad argument) in function 'cvtColor'

Overload resolution failed:

• src is not a numpy array, neither a scalar
• Expected Ptrcv::UMat for argument 'src'

1 Frames collected

am i doing something wrongly?

[glenn-jocher]

Hello @codinglearningnovice,

Thank you for your update. It looks like the error you're encountering is related to the cvtColor function from OpenCV, which expects a numpy array but is receiving a different type.

To help us investigate further, could you please provide a minimum reproducible example of your code? This will allow us to reproduce the issue on our end and find a solution more effectively. You can find guidelines on how to create one here.

In the meantime, let's ensure that the person_masks variable is indeed a numpy array before passing it to cvtColor. Here’s a revised snippet that includes a check:

while(count < TRAIN_SIZE):
    try:
        ret, frame = cap.read()

        if currentFrame % FRAME_SKIP == 0:
            count += 1
            if count % int(TRAIN_SIZE/10) == 0:
                print(str((count/TRAIN_SIZE)*100) + "% done")

            # Perform human segmentation
            results = model(frame)

            for result in results:
                person_masks = result.masks[result.boxes.cls == 0].numpy()  # Ensure masks are numpy arrays
                person_mask_3ch = cv2.cvtColor(person_masks, cv2.COLOR_GRAY2BGR)
                masked_frame = cv2.bitwise_and(frame, person_mask_3ch)

                inverted_mask = cv2.bitwise_not(person_mask_3ch)

                result_frame = cv2.bitwise_and(masked_frame, inverted_mask)

                resized_frame = cv2.resize(result_frame, (output_width, output_height))

                name = 'trydata/resized_frame.jpg' + str(count) + '.jpg'

                cv2.imwrite(name, resized_frame)

                video.write(resized_frame.astype('uint8'))

    except Exception as e:
        print(e)
        break

    currentFrame += 1

print(str(count) + " Frames collected")
cap.release()
video.release()

Additionally, please ensure you are using the latest versions of torch and ultralytics. You can upgrade them using the following commands:

pip install --upgrade torch ultralytics

If the issue persists, please share the minimum reproducible example so we can assist you further. Thank you for your cooperation! 😊

[codinglearningnovice]

Bug description:

When running inference on a video to segment the person and manipulate each frame, I get an error related to the expected input from the cv2.cvt, seems to be a type mismatch

MRE:

import cv2
from ultralytics import YOLO

# Load the YOLOv8 segmentation model
model = YOLO("yolov8n-seg.pt")

cap = cv2.VideoCapture('dancee.mp4')
output_width, output_height = 96, 64  # Adjust as needed
video = cv2.VideoWriter('output_video.mp4', cv2.VideoWriter_fourcc(*'mp4v'), 30, (output_width, output_height))


            # Perform human segmentation
            results = model(frame)

            for result in results:
                person_masks = result.masks[result.boxes.cls == 0].numpy()  # Ensure masks are numpy arrays
                person_mask_3ch = cv2.cvtColor(person_masks, cv2.COLOR_GRAY2BGR)
                masked_frame = cv2.bitwise_and(frame, person_mask_3ch)

                inverted_mask = cv2.bitwise_not(person_mask_3ch)

                result_frame = cv2.bitwise_and(masked_frame, inverted_mask)

                resized_frame = cv2.resize(result_frame, (output_width, output_height))

                name = 'trydata/resized_frame.jpg' + str(count) + '.jpg'

                cv2.imwrite(name, resized_frame)

                video.write(resized_frame.astype('uint8'))

    except Exception as e:
        print(e)
        break

    currentFrame += 1

print(str(count) + " Frames collected")
cap.release()
video.release()

Error message:

OpenCV(4.8.0) 👎 error: (-5:Bad argument) in function 'cvtColor'

Overload resolution failed:

• src is not a numpy array, neither a scalar
• Expected Ptrcv::UMat for argument 'src'

Dependencies:

ultralytics==8.2.0

[glenn-jocher]

Hello @codinglearningnovice,

Thank you for providing a detailed description of the issue and the minimum reproducible example (MRE). It looks like the error is due to a type mismatch when using cv2.cvtColor. Let's ensure that the person_masks variable is indeed a numpy array before passing it to cv2.cvtColor.

First, please make sure you are using the latest versions of torch and ultralytics. You can upgrade them using the following commands:

pip install --upgrade torch ultralytics

Here’s a revised version of your code snippet that includes a check to ensure person_masks is a numpy array:

import cv2
from ultralytics import YOLO

# Load the YOLOv8 segmentation model
model = YOLO("yolov8n-seg.pt")

cap = cv2.VideoCapture('dancee.mp4')
output_width, output_height = 96, 64  # Adjust as needed
video = cv2.VideoWriter('output_video.mp4', cv2.VideoWriter_fourcc(*'mp4v'), 30, (output_width, output_height))

count = 0
TRAIN_SIZE = 1000  # Adjust as needed
FRAME_SKIP = 5  # Adjust as needed
currentFrame = 0

while count < TRAIN_SIZE:
    try:
        ret, frame = cap.read()
        if not ret:
            break

        if currentFrame % FRAME_SKIP == 0:
            count += 1
            if count % int(TRAIN_SIZE / 10) == 0:
                print(f"{(count / TRAIN_SIZE) * 100}% done")

            # Perform human segmentation
            results = model(frame)

            for result in results:
                person_masks = result.masks[result.boxes.cls == 0].numpy()  # Ensure masks are numpy arrays
                if person_masks.size == 0:
                    continue  # Skip if no person masks are found

                person_mask_3ch = cv2.cvtColor(person_masks[0], cv2.COLOR_GRAY2BGR)  # Convert the first mask to 3 channels
                masked_frame = cv2.bitwise_and(frame, person_mask_3ch)

                inverted_mask = cv2.bitwise_not(person_mask_3ch)

                result_frame = cv2.bitwise_and(masked_frame, inverted_mask)

                resized_frame = cv2.resize(result_frame, (output_width, output_height))

                name = f'trydata/resized_frame_{count}.jpg'

                cv2.imwrite(name, resized_frame)

                video.write(resized_frame.astype('uint8'))

    except Exception as e:
        print(e)
        break

    currentFrame += 1

print(f"{count} Frames collected")
cap.release()
video.release()

This code ensures that person_masks is a numpy array and handles cases where no person masks are found. Additionally, it converts the first mask to 3 channels before applying cv2.cvtColor.

If the issue persists, please provide any additional details or errors you encounter. This will help us further investigate and provide a more accurate solution.

Thank you for your patience and cooperation! 😊