* @attention Vector size should be given as a preprocessor argument using -DVEC_SIZE=size. e.g. -DVEC_SIZE=16
* @attention Data type can be passed using the -DDATA_TYPE compile flag, e.g. -DDATA_TYPE=float
*
- * @param[in] in_ptr Pointer to the source image. Supported data types: U8.
+ * @param[in] in_ptr Pointer to the source image. Supported data types: F16/F32.
* @param[in] in_stride_x Stride of the source image in X dimension (in bytes)
* @param[in] in_step_x in_stride_x * number of elements along X processed per work item (in bytes)
* @param[in] in_offset_first_element_in_bytes Offset of the first element in the source image
- * @param[out] out_ptr Pointer to the destination image. Supported data types: U8.
+ * @param[out] out_ptr Pointer to the destination image. Supported data types: same as @p input_ptr
* @param[in] out_stride_x Stride of the destination image in X dimension (in bytes)
* @param[in] out_step_x out_stride_x * number of elements along X processed per work item (in bytes)
* @param[in] out_offset_first_element_in_bytes Offset of the first element in the destination image