Batch Labeling Videos in CoLab

[Kagiro-K]

I had a really basic question to ask as I am a complete novice at coding. I am utilizing your CoLab notebook for the "Predicting and tracking instances in uploaded video" part. I am trying to batch process a large number of clips I've made but was having difficulty writing the correct segment of code for that.
I would love any feedback you can provide!

Here is one example of what I've tried:

import os

# Specify the folder path containing the videos
folder_path = "/content/drive/My Drive/SLEAP/NewSkeleClipped/clips"

# Iterate over the files in the folder
for file_name in os.listdir(folder_path):
    # Construct the full file path
    file_path = os.path.join(folder_path, file_name)
    
    # Check if the item is a file and ends with a video extension
    if os.path.isfile(file_path) and file_name.endswith((".mp4")):
        # Run sleap-track on the video
      !sleap-track {file_path} \
              --frames 0-208\
              --tracking.tracker simple \
              -m models/single_instance_MasterFile

And this is the error I get afterward:

Would appreciate any guidance on what the error is referring to and how to remedy it.

Thank you!

[roomrys]

Hi @Kagiro-K,

SLEAP uses the from_filename method from the Video class to determine what the backend type should be (i.e. MediaVideo for mp4/avi). We should really also print the "specified filename" in the ValueError.... My guess would be that your filename might have a space which would then be parsed in as multiple arguments.

Can you try putting quotes around the {file_path} s.t. you call sleap-track "{file_path}". If spaces are the problem, then this will ensure that the spaces do not separate your filepath into multiple arguments.

Can you also add a try, except block around the sleap-track command to print the filename (which we should be printing for you in the ValueError, but are not at the moment):

try:
    !sleap-track "{file_path}"  # Add the rest of your command
except Exception as e:
    print(f"filepath: {file_path}")
    raise e

This will let us know what the file_path is (only if the command errors out) and give us some insight into why it is causing problems.

Let us know how it goes.

Thanks,
Liezl

[Kagiro-K]

Thank you so much for your response! Adding the quotations seemed to do the trick! It's unusual because my file names did not have any spaces. But I also moved the trained model into the same folder as the videos, instead of having those files in a subfolder. This will greatly speed things up in CoLab where I can train, update labels locally, and continue to improve the model. Thank you! I really appreciate the speedy response!

…

________________________________ From: Liezl Maree ***@***.***> Sent: Tuesday, June 20, 2023 5:50 PM To: talmolab/sleap ***@***.***> Cc: Harmain Rafi ***@***.***>; Mention ***@***.***> Subject: Re: [talmolab/sleap] Batch Labeling Videos in CoLab (Discussion #1356) Hi @Kagiro-K<https://urldefense.com/v3/__https://github.com/Kagiro-K__;!!IaT_gp1N!3lRFlQYZl6ZaYVVRttDUii5w3O-2_Fzg0X93U3rcJy4HgpOKJRrENqJNXjHqQQDp5QJOCGFCLsKKlCgND3NAeJqVfnzVlw$>, SLEAP uses the from_filename method from the Video class to determine what the backend type should be (i.e. MediaVideo for mp4/avi). We should really also print the "specified filename" in the ValueError.... My guess would be that your filename might have a space which would then be parsed in as multiple arguments. Can you try putting quotes around the {file_path} s.t. you call sleap-track "{file_path}". If spaces are the problem, then this will ensure that the spaces do not separate your filepath into multiple arguments. Can you also add a try, except block around the sleap-track command to print the filename (which we should be printing for you in the ValueError, but are not at the moment): try: !sleap-track "{file_path}" # Add the rest of your command except Exception as e: print(f"filepath: {file_path}") raise e This will let us know what the file_path is (only if the command errors out) and give us some insight into why it is causing problems. Let us know how it goes. Thanks, Liezl — Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/talmolab/sleap/discussions/1356*discussioncomment-6235526__;Iw!!IaT_gp1N!3lRFlQYZl6ZaYVVRttDUii5w3O-2_Fzg0X93U3rcJy4HgpOKJRrENqJNXjHqQQDp5QJOCGFCLsKKlCgND3NAeJoUxSdRKQ$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/A6IXUSPCRC5VO424FNDKH43XMILLFANCNFSM6AAAAAAZJ5NKPM__;!!IaT_gp1N!3lRFlQYZl6ZaYVVRttDUii5w3O-2_Fzg0X93U3rcJy4HgpOKJRrENqJNXjHqQQDp5QJOCGFCLsKKlCgND3NAeJotb8-sOw$>. You are receiving this because you were mentioned.Message ID: ***@***.***>