Janusz Lisiecki

Hi @DatCanCode, There is no direct OpenCV `copyMakeBorder` counterpart in DALI. What you can try is: - check [the paste operator](https://docs.nvidia.com/deeplearning/dali/user-guide/docs/supported_ops.html#nvidia.dali.fn.paste) - write your own operator using [this guide](https://docs.nvidia.com/deeplearning/dali/user-guide/docs/examples/custom_operations/index.html)

Hi @wangchengdng, If the constant border is what you are looking for you can use the slice operator for it: ``` images_pad = fn.slice(images, (-0.3, -0.3), (1.6, 1.6), out_of_bounds_policy="pad", normalized_shape=True,...

Hi @zhaoyang-0204, JAX is still advertised as an experimental project and we haven't investigated how to integrate DALI with it. Thus there is no recommended way of using DALI with...

Hi @neurochen, You may check [the relevant test file](https://github.com/NVIDIA/DALI/blob/main/dali/test/python/test_operator_get_property.py) for this operator to get the gist of using it.

Hi @neurochen, Could you be more specific regarding: > the sorting rules for classes? Do you want to know how DALI sorts samples?

Hi @taohui, Please check https://github.com/NVIDIA/DALI/issues/2789 for reference. You may also read https://github.com/NVIDIA/DALI/issues/3813 to learn more about other differences in the pipelines that are most commonly used in image processing.

Hi @MattWittbrodt, Thank you for reporting the problem. I managed to reproduce it as well. We will keep you posted.

The minimal repro I'm using is (just for the reference): ``` from nvidia.dali import fn, pipeline_def # Pipe parameters batch_size=1 sequence_length = 355 @pipeline_def def video_pipe(): video, label = \...

Tracked as nv3452920.

I don't see anything obviously wrong in the way how DALI uses the Video SDK. I just asked the relevant team to check on their side. It may take a...