NGC | Catalog
Welcome Guest
CatalogContainersMaxine Video Effects SDK Beta v0.6

Maxine Video Effects SDK Beta v0.6

For pull tags and more information, please view on a desktop device.
Logo for Maxine Video Effects SDK Beta v0.6

Description

Maxine Linux Video Effects SDK Public Beta container version 0.6

Publisher

NVIDIA

Latest Tag

0.6.5

Modified

June 28, 2022

Compressed Size

6.59 GB

Multinode Support

No

Multi-Arch Support

No

0.6.5 (Latest) Scan Results

Linux / amd64

Video Effects SDK

The VFX SDK provides five features that can be organized into three categories. Super Resolution and Upscaler enhance the video resolution quality and look and feel of the video streams. Artifact Reduction and Video Noise Removal remove video noise from the streams. Artifact Reduction reduces video compression artifacts, and Video Noise Removal reduces the video degradation caused by low light conditions or a lower quality camera. Lastly, AI Green Screen gives semantic segmentation tailored to typical video conferencing environments.

Prerequisites

There are three main pre-requisites for VFX SDK's containers:

  • NVIDIA Drivers NVIDIA Drivers (465.19.01+ for recommended for Maxine) are required to use NVIDIA GPUs here. Note: For older drivers please review CUDA Backward Compatibilty(experience may or may not be optimal).
  • Docker (19.03+) and the latest version of NVIDIA-Docker
  • NGC API Key for logging to NVIDIA's registry. Details are available here.

Running the container

Use the following commands to run the container.

# (optional) use if using a remote system and want to view the output on screen
xhost +

# Add "-e DISPLAY=$DISPLAY" for remote display
docker run --gpus all -it --rm -v : nvcr.io/nvidia/maxine/maxine-videoeffects-sdk:

If using depricated nvidia-docker, based on a version of docker prior to 19.0, use:

nvidia-docker run -it --rm -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -w nvcr.io/nvidia/maxine/maxine-videoeffects-sdk:

Note that the above command mounts the host's X11 display in the guest filesystem to render output videos.

Documentation & resources

There are a plethora of resources you can access like our Programming Guide, Getting Started Guide, Devloper Blogs and API Videos. All these resoureces are accessible via our Getting Started Page.

The VFX SDK's programming guide can be found within the container at /usr/local/VideoFX/share/docs and here. This guide contains information about:

  • Supplementary feature information and information for selecting the most suitable configurations for specific use-cases
  • Detailed VFX API descriptions with best practices and examples
  • Information for using sample applications
  • Bare-metal installation requirements

The container houses sample applications for feature demonstrations as well as an implementation example. Check the "Running a sample application" section for more details.

The sample applications provide an example for using Maxine-specific APIs. Some familiarity with CUDA is required. To brush up on CUDA basics, check out the tutorials, the documentation, and the best practices guide.

Running a sample application

To use the sample applications, we need to build them first.

Use the following command to build the aforementioned sample applications. By default, the application is built in /root/mysamples.

bash /usr/local/VideoFX/share/build_samples.sh

Once the applications are built, navigate to /root/mysamples to use them. In this folder, multiple run_.sh files are generated for running the sample applications.

cd /root/mysamples
bash run_videoeffects.sh

Let's take a closer look at a section of run_videoeffects.sh

#!/bin/sh

. ./setup_env.sh

VideoEffectsApp \
 --model_dir=$_VFX_MODELS \
 --in_file=$_VFX_SHARE/samples/input/input1.jpg \
 --out_file=ar_1.png \
 --effect=ArtifactReduction \
 --strength=1 \
 --show

...

Here, the user can select the effect with --effect, specify the I/O with --in_file and --out_file, choose the model with --strength. Please explore Section 2 of the programming guide for more details. You can find a walkthrough of the Video Effects SDK API here on YouTube

Changelog

  • AI Green Screen filter
  • Quality mode is updated for higher quality
  • Added CUDA graph optimization which may improve overall latency under a GPU-intensive workload.
  • Super Resolution filter - 4K input support for 1.33x, 1.5x and 2x super resolution
  • Migrated to TensorRT 8.0.1.6
  • Migrated to CUDA 11.3u1 and cuDNN 8.2

License

The AFX SDK license can be found within the container and on our Getting Started Page

Ethical AI

NVIDIA's platforms and application frameworks enable developers to build a wide array of AI applications. Please, consider potential algorithmic bias when choosing or creating the models being deployed. Work with the model's developer to ensure:

  • The model meets the requirements for the relevant industry and use case
  • The necessary instruction and documentation are provided to understand error rates, confidence intervals, and results
  • The model is being used under the conditions and in the manner intended.