Dynamic heterogeneous task processing

A non-transitory computer-readable storage medium comprising instructions stored thereon that, when executed by at least one processor, are configured to cause a computing system to:

• process a stream of data in a sequence of tasks, the processing including:
• receiving a first block of data of the stream of data;
• determining features associated with the first block of data;
• selecting, based on the features, one of a first task to process the first block of data or a second task to process the first block of data;
• and
• in response to selecting the second task, shift an output of the second task in time to align the output of the second task with a predicted output of the first task processing a second block of data of the stream of data.

An apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to

• process a stream of data in a sequence of neural networks, the processing including:
• receiving a first block of data of the stream of data;
• determining features associated with the first block of data;
• selecting, based on the features, one of a first a neural network to process the first block of data or a second neural network to process the first block of data;
• and
• in response to selecting the second neural network, shift an output of the second neural network in time to align the output of the second neural network with a predicted output of the first neural network processing a second block of data of the stream of data.