Skip to content
AI News HubLIVE
Original source2 min read

SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness

Summary

SCULPT is a training-time method that improves post-training quantization (PTQ) readiness during ordinary FP32 fine-tuning by suppressing quantization-hostile activation skewness and kurtosis and learning exportable percentile-based clipping bounds, enabling INT8 and W4A8 deployment without QAT or runtime activation transforms.

SourcearXiv Computer VisionAuthor: Bharadwaj Kavuri, Sourav Babu-PK, Varadhraj Ellapan, Pullarao Maddu, Prasad Deshpande
SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 1 Sep 2026]

Title:SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness

View a PDF of the paper titled SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness, by Bharadwaj Kavuri and 4 other authors

View PDF HTML (experimental)

Abstract:Edge vision models are difficult to deploy on resource-constrained hardware, making low-bit post-training quantization (PTQ) attractive. In practice, standard FP32 training often produces heavy-tailed activation distributions whose outliers destabilize activation quantization: preserving the full range wastes quantization bins on rare extremes, while aggressive clipping causes information loss. Existing solutions typically rely on quantization-aware training (QAT), which adds training complexity and bit-width coupling, or advanced PTQ procedures that repair the model after training.

We present SCULPT (Statistical Clipping and Uniform Loss for Post-Training), a training-time method that improves PTQ readiness during ordinary FP32 fine-tuning. SCULPT combines a topology-aware activation regularizer that suppresses quantization-hostile skewness and kurtosis with a stable percentile-based clipping mechanism that learns deployment-ready activation bounds. Unlike QAT, SCULPT does not simulate quantization during optimization; unlike post hoc outlier-repair PTQ methods, it does not require runtime activation transformations. The learned clipping bounds can be exported directly into a standard PTQ workflow for low-bit deployment, including INT8 and lower-bit settings such as W4A8.

Comments: Published at Irish Machine Vision and Image Processing Conference (IMVIP) 2026. Paper 043. 8 pages, 3 figures

Subjects:

Computer Vision and Pattern Recognition (cs.CV)

Cite as: arXiv:2609.01743 [cs.CV]

(or arXiv:2609.01743v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2609.01743

arXiv-issued DOI via DataCite

Journal reference: Irish Machine Vision and Image Processing Conference (IMVIP), 2026

Submission history

From: Prasad Deshpande [view email] [v1] Tue, 1 Sep 2026 18:10:19 UTC (352 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness, by Bharadwaj Kavuri and 4 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CV

new | recent | 2026-09

Change to browse by:

cs

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • SCULPT regularizes activation distributions during FP32 training to reduce quantization-hostile skewness and kurtosis.
  • It learns stable percentile-based clipping bounds that export directly into standard PTQ workflows.
  • Unlike QAT, it does not simulate quantization during optimization; unlike post hoc PTQ repair, it needs no runtime activation transforms.
  • Published at IMVIP 2026, it targets edge vision deployment at INT8 and lower-bit settings such as W4A8.

Highlights and analysis are generated automatically and may contain errors. Check the original source.