run() [1/2]

VisionResult MayaFlux::Yantra::VisionGpuExecutor::run	(	const Kinesis::Vision::VisionSequence &	sequence,
		const std::shared_ptr< Core::VKImage > &	image,
		uint32_t	w,
		uint32_t	h
	)

Execute a VisionSequence on the GPU using this instance's own lazily-constructed context set.

m_contexts is built on first call and reused for every subsequent call to this overload on the same VisionGpuExecutor instance. The primary entry point; prefer this over the explicit-contexts overload unless a caller specifically needs to inspect or share a VisionGpuContexts across multiple calls outside this class.

Parameters

sequence	Ordered steps to execute.
image	Input frame in eShaderReadOnlyOptimal layout.
w	Frame width in pixels.
h	Frame height in pixels.

Returns

VisionResult matching the VisionExecutor::run() contract.

Definition at line 1323 of file VisionGpuDispatch.cpp.

{
    if (!m_contexts)
        m_contexts = std::make_unique<VisionGpuContexts>();
 
    return run(*m_contexts, sequence, image, w, h);
}

References h, image, and run().

Here is the call graph for this function: