Expand description
Batch processing utilities for the OCR pipeline.
This module provides structures and functions for handling batched data in the OCR pipeline, including batching of input data, sampling, and tensor operations for batched processing.
Modules§
- dynamic
- Dynamic batching for OCR pipeline components.
Structs§
- Batch
Data - Data structure for holding batched input data.
- Batch
Sampler - A sampler that creates batches of data with a specified batch size.
- ToBatch
- A struct for converting image data into batched tensor format.
Type Aliases§
- Tensor1D
- A 1-dimensional tensor represented as a dynamic-dimensional array of f32 values.
- Tensor2D
- A 2-dimensional tensor represented as a 2D array of f32 values.
- Tensor3D
- A 3-dimensional tensor represented as a 3D array of f32 values.
- Tensor4D
- A 4-dimensional tensor represented as a 4D array of f32 values.