Back to Search Document details
41st Meeting: by teleconference, CH, January 2026 2026-01-17 01:43
AhG11 Overflow aware neural-network model quantization
Authors: Elena Alshina Huawei Timofey Solovyev Huawei
Abstract
This contribution provides details of overflow aware neural network quantizer which guarantees bit-exact behaviour and so device interoperability. Additionally to quantizing neural network weights and activations to integer values, the sum of model parameters’ absolute values must not exceed ration between max values of accumulation register and temporal buffer. It is recommended to conduct theoretical analysis for dynamic range of neural-network based algorithms and provide mathematical proof ensuring bit-exact behaviour, as a condition for adoption to the standard draft (if any neural-network based tool is considered).
Meeting Notes
JVET-AO0148 AhG11 Overflow aware neural-network model quantization [E. Alshina, T. Solovyev (Huawei)]

This contribution provides details of overflow aware neural network quantizer which guarantees bit-exact behaviour and so device interoperability. Additionally to quantizing to integer values neural network the sum of model parameters’ absolute values must not exceed ration between max values of accumulation register and temporal buffer. It was recommended to conduct theoretical analysis for dynamic range of neural-network based algorithms and provide mathematical proof ensure bit-exact behaviour as condition for adoption to the standard draft (if any neural-network based tool is considered).

Notes were recorded by F. Galpin.

Similar methodology to what is done in legacy video codec design is used: design the operations for the worst case inputs in term of accumulation/overflow. It is part of the training process.

The proponent was asked for testing the training integration on current NNVC tools. Currently, this was not tried but believed to be possible with NNLF tools.

The proponent recommended to conduct theoretical analysis for dynamic range of neural-network based algorithms and provide mathematical proof ensure bit-exact behaviour as condition for adoption to the standard draft.

ARCHIVE :
JVET-AO0148_v3.docx JVET-AO0148_v3_clean.docx JVET-AO0148_v4.pptx
Decisions
adopted
The proponent recommended to conduct theoretical analysis for dynamic range of neural-network based algorithms and provide mathematical proof ensure bit-exact behaviour as condition for adoption to the standard draft.
Citation