For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo module

bicubic

This module provides CPU and GPU implementations for bicubic interpolation.

Bicubic interpolation is a 2D extension of cubic interpolation for resampling digital images. It uses the weighted average of the 4x4 neighborhood of pixels around the target location to compute the interpolated value.

Functions

• ​cpu_bicubic_kernel: Perform bicubic interpolation on a TileTensor of form NCHW.
• ​cubic_kernel: Cubic interpolation kernel matching PyTorch/torchvision's BICUBIC filter.
• ​gpu_bicubic_kernel: Perform bicubic interpolation using GPU.
• ​map_output_to_input_coord: Map output pixel coordinate to input coordinate using center alignment. This implements the standard coordinate mapping for image resizing: input_coord = (output_coord + 0.5) * scale - 0.5 The +0.5 and -0.5 terms ensure pixel centers are aligned properly. Args: output_coord: Output pixel coordinate. scale: Scale factor (input_size / output_size). Returns: Corresponding input coordinate as a float.
• ​resize_bicubic: Perform bicubic interpolation.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit