Reducing latency in video encoding and decoding
Techniques and tools for reducing latency in video encoding and decoding by constraining latency due to reordering of video frames, and by indicating the constraint on frame reordering latency with one or more syntax elements that accompany encoded data for the video frames. For example, a real-time communication tool with a video encoder sets a syntax element that indicates a constraint on frame reordering latency, which is consistent with inter-frame dependencies between multiple frames of a video sequence, then outputs the syntax element. A corresponding real-time communication tool with a video decoder receives the syntax element that indicates the constraint on frame reordering latency, determines the constraint on frame reordering latency based on the syntax element, and uses the constraint on frame reordering latency to determine when a reconstructed frame is ready for output (in terms of output order).
1 . One or more non-transitory computer-readable media having stored thereon computer-executable instructions for causing one or more processing units, when programmed thereby, to perform operations comprising:
receiving syntax elements that indicate different constraints on frame reordering latency for different temporal layers of a temporal hierarchy;
determining, as temporal resolution for output, one of the different temporal layers of the temporal hierarchy;
for the determined layer, selecting one of the different constraints on frame reordering latency, the selected constraint on frame reordering latency being determined, relative to a maximum size of frame memory for reordering, using a given one of the syntax elements that indicate the different constraints on frame reordering latency;
receiving encoded data for a given frame;
decoding the encoded data for the given frame to reconstruct the given frame; and
outputting the reconstructed frame, wherein the selected constraint on frame reordering latency is used to determine when the reconstructed frame is ready for output.
2 . The one or more non-transitory computer-readable media of claim 1 , wherein the given one of the syntax elements indicates an integer count for the selected constraint on frame reordering latency relative to the maximum size of frame memory for reordering.
3 . The one or more non-transitory computer-readable media of claim 2 , wherein the selected constraint on frame reordering latency is determined as the maximum size of frame memory for reordering plus the integer count for the selected constraint on frame reordering latency minus 1.
4 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:
receiving a syntax element that indicates the maximum size of frame memory for reordering, the syntax element that indicates the maximum size of frame memory for reordering being different than the syntax elements that indicate the different constraints on frame reordering latency.
5 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:
receiving a syntax element that indicates a maximum size of a decoded picture buffer, the syntax element that indicates the maximum size of the decoded picture buffer being different than the syntax elements that indicate the different constraints on frame reordering latency.
6 . The one or more non-transitory computer-readable media of claim 1 , wherein the syntax elements that indicate the different constraints on frame reordering latency are signaled as part of a sequence parameter set.
7 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:
receiving a flag that indicates presence or absence of the syntax elements that indicate the different constraints on frame reordering latency.
8 . The one or more non-transitory computer-readable media of claim 1 , wherein the different temporal layers of the temporal hierarchy include a first layer, a second layer, and a third layer, wherein the second layer doubles temporal resolution of the first layer, and wherein the third layer doubles temporal resolution of the second layer.
9 . In a computer system that implements a video encoder, a method comprising:
setting syntax elements that indicate different constraints on frame reordering latency for different temporal layers of a temporal hierarchy;
outputting the syntax elements that indicate the different constraints on frame reordering latency, thereby facilitating determination of when a reconstructed frame is ready for output according to a selected constraint on frame reordering latency that is determined, relative to a maximum size of frame memory for reordering, using a given one of the syntax elements that indicate the different constraints on frame reordering latency;
encoding a given frame, thereby producing encoded data for the given frame, wherein the encoding uses inter-frame dependencies that are consistent with the different constraints on frame reordering latency; and
outputting the encoded data for the given frame.
10 . The method of claim 9 , wherein the given one of the syntax elements indicates an integer count for a given one of the different constraints on frame reordering relative to the maximum size of frame memory for reordering.
11 . The method of claim 10 , wherein the given one of the different constraints on frame reordering latency is determined as the maximum size of frame memory for reordering plus the integer count for the given one of the different constraints on frame reordering latency minus 1.
12 . The method of claim 9 , further comprising:
setting a syntax element that indicates the maximum size of frame memory for reordering, the syntax element that indicates the maximum size of frame memory for reordering being different than the syntax elements that indicate the different constraints on frame reordering latency; and
outputting the syntax element that indicates the maximum size of frame memory for reordering.
13 . The method of claim 9 , further comprising:
setting a syntax element that indicates a maximum size of a decoded picture buffer, the syntax element that indicates the maximum size of the decoded picture buffer being different than the syntax elements that indicate the different constraints on frame reordering latency; and
outputting the syntax element that indicates maximum size of the decoded picture buffer.
14 . The method of claim 9 , wherein the syntax elements that indicate the different constraints on frame reordering latency are signaled as part of a sequence parameter set.
15 . The method of claim 9 , further comprising:
setting a flag that indicates presence or absence of the syntax elements that indicate the different constraints on frame reordering latency; and
outputting the flag that indicates presence or absence of the syntax elements that indicate the different constraints on frame reordering latency.
16 . The method of claim 9 , wherein the different temporal layers of the temporal hierarchy include a first layer, a second layer, and a third layer, wherein the second layer doubles temporal resolution of the first layer, and wherein the third layer doubles temporal resolution of the second layer.
17 . One or more non-transitory computer-readable media having programmed thereon a flag that at least in part indicates presence or absence of syntax elements that indicate different constraints on frame reordering latency for different temporal layers of a temporal hierarchy, the one or more computer-readable media further having programmed thereon the syntax elements that indicate the different constraints on frame reordering latency as well as encoded data for a given frame, the flag, the syntax elements, and the encoded data being usable to cause a video decoder, when processing the flag, the syntax elements, and the encoded data in a computer system having one or more processing units and memory, to perform operations that include:
based at least in part on the flag, determining that the syntax elements that indicate the different constraints on frame reordering latency are present;
determining, as temporal resolution for output, one of the different temporal layers of the temporal hierarchy;
for the determined layer, selecting one of the different constraints on frame reordering latency, selected constraint on frame reordering latency being determined, relative to a maximum size of frame memory for reordering, using a given one of the syntax elements that indicate the different constraints on frame reordering latency;
based on the encoded data for the given frame, decoding the encoded data for the given frame to reconstruct the given frame; and
outputting the reconstructed frame, wherein the selected constraint on frame reordering latency is used to determine when the reconstructed frame is ready for output.
18 . The one or more non-transitory computer-readable media of claim 17 , further having stored thereon a syntax element that indicates the maximum size of frame memory for reordering, the syntax element that indicates the maximum size of frame memory for reordering being different than the syntax elements that indicate the different constraints on frame reordering latency.
19 . The one or more non-transitory computer-readable media of claim 17 , further having stored thereon a syntax element that indicates a maximum size of a decoded picture buffer, the syntax element that indicates the maximum size of the decoded picture buffer being different than the syntax elements that indicate the different constraints on frame reordering latency.