[data request] LVIS

[netw0rkf10w]

• Name of dataset: LVIS
• URL of dataset: https://www.lvisdataset.org
• License of dataset: Creative Commons Attribution 4.0 License
• Short description of dataset and use case(s): A new dataset for Large Vocabulary Instance Segmentation. There will be ∼2 million high-quality instance segmentation masks for over 1000 entry-level object categories in 164k images.

Folks who would also like to see this dataset in tensorflow/datasets, please thumbs-up so the developers can know which requests to prioritize.

And if you'd like to contribute the dataset (thank you!), see our guide to adding a dataset.

[Player0109]

I am a deep learning enthusiast, so I want to work on this dataset. Can I do so? @netw0rkf10w

[Marvin182]

LVIS is now included in TFDS. github.com/tensorflow/datasets/pull/2836

[Conchylicultor]

Thanks Marvin for your contribution! Closing this

[XinyueZ]

How to load this dataset, I still cannot see it in the tfds..... ? Fire errors with following, even tried !pip install tfds-nightly before running all imports. 😅

train_data, info = tfds.load('lvis', split='train', as_supervised=True, with_info=True)
validation_data = tfds.load('lvis', split='validation', as_supervised=True)
test_data = tfds.load('lvis', split='test', as_supervised=True)

[Conchylicultor]

What is the stacktrace ?

[XinyueZ]

What is the stacktrace ?

@Conchylicultor

---------------------------------------------------------------------------
NotFoundError                             Traceback (most recent call last)
/usr/local/lib/python3.7/dist-packages/tensorflow_datasets/core/utils/py_utils.py in try_reraise(*args, **kwargs)
    391   try:
--> 392     yield
    393   except Exception as e:  # pylint: disable=broad-except

15 frames
NotFoundError: /usr/local/lib/python3.7/dist-packages/tensorflow_datasets/object_detection/lvis/lvis_classes.txt; No such file or directory

The above exception was the direct cause of the following exception:

RuntimeError                              Traceback (most recent call last)
/usr/local/lib/python3.7/dist-packages/tensorflow_datasets/core/utils/py_utils.py in reraise(e, prefix, suffix)
    359     else:
    360       exception = RuntimeError(f'{type(e).__name__}: {msg}')
--> 361     raise exception from e
    362   # Otherwise, modify the exception in-place
    363   elif len(e.args) <= 1:

RuntimeError: NotFoundError: Failed to construct dataset lvis: /usr/local/lib/python3.7/dist-packages/tensorflow_datasets/object_detection/lvis/lvis_classes.txt; No such file or directory

Full code

import os
import tensorflow as tf

!pip install tfds-nightly
import tensorflow_datasets as tfds

train_data, info = tfds.load('lvis', split='train', as_supervised=True, with_info=True)
validation_data = tfds.load('lvis', split='validation', as_supervised=True)
test_data = tfds.load('lvis', split='test', as_supervised=True)

Or my approach is not correct ?