ClikaRT::io::FrameBatch

struct

Header: ClikaRT/io/io.h

A decoded batch of frames plus the alignment metadata a video model needs: the packed [K, H, W, 3] UInt8 frames, the source frame indices they were taken from, and per-frame pts_seconds / duration_seconds. Indexable / iterable: operator[] yields a single-frame view, and begin()/end() iterate single-frame views in order.

Member functions

size()

std::size_t size() const noexcept

Number of frames in the batch.

Declared in ClikaRT/io/io.h, line 305

empty()

bool empty() const noexcept

True when the batch holds no frames.

Declared in ClikaRT/io/io.h, line 307

Data members

frames

Tensor frames

[K, H, W, 3] UInt8 (host)

Declared in ClikaRT/io/io.h, line 296

indices

std::vector<std::int64_t> indices

the source frame indices, in order

Declared in ClikaRT/io/io.h, line 297

pts_seconds

std::vector<double> pts_seconds

The container's presentation timestamp of each frame, in seconds: the clip's own timeline, offsets and uneven spacing included, not the frame index divided by the frame rate.

Declared in ClikaRT/io/io.h, line 301

duration_seconds

std::vector<double> duration_seconds

display duration of each frame

Declared in ClikaRT/io/io.h, line 302