Tlt-infer - input image

Accelerated Computing Intelligent Video Analytics TAO Toolkit
1

When I give the command line argument for the images, do I need to resize them before hand or will it automatically be taken care of the match the input size of the model?

2

If you want to run tlt-train at Detection network, it is necessary to resize the images/labels offline.
See tlt user guide for more details.

Object Detection

DetectNet_v2

Input size: C * W * H (where C = 1 or 3, W > =480, H >=272 and W, H are multiples of 16)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

FasterRCNN

Input size: C * W * H (where C = 1 or 3; W > =160; H >=160)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

SSD

Input size: C * W * H (where C = 1 or 3, W >= 128, H >= 128, W, H are multiples of 32)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

DSSD

Input size: C * W * H (where C = 1 or 3, W >= 128, H >= 128, W, H are multiples of 32)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

YOLOv3

Input size: C * W * H (where C = 1 or 3, W >= 128, H >= 128, W, H are multiples of 32)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

RetinaNet

Input size: C * W * H (where C = 1 or 3, W >= 128, H >= 128, W, H are multiples of 32)
Image format: JPG, JPEG, PNG
Label format: KITTI detection

Note: The tlt-train tool does not support training on images of multiple resolutions, or resizing images during training. All of the images must be resized offline to the final training size and the corresponding bounding boxes must be scaled accordingly.

1 Like
3

For tlt-infer, it is not needed to resize.