FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers

• URL: http://arxiv.org/abs/2609.01683v1
• Date: Tue, 01 Sep 2026 13:59:28 GMT
• Title: FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers
• Abstract summary: We present a forward-only test-time adaptation (TTA) method that operates on deployed, BN-folded, integer-only convolutional networks.<n>We validate bit-exact int8 convolution execution and deploy on an ESP32-S3 microcontroller.
• Score: 0.3499870393443268
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision models deployed on microcontrollers (MCUs) are quantized to integer-only arithmetic and run in inference-only runtimes that do not carry the machinery backpropagation needs: the standard tool for adapting a model to the distribution shift (sensor noise, blur, lighting) it meets in the field. Existing forward-only test-time adaptation (TTA) methods either run only on server- or edge-GPU-class models (not true microcontroller integer execution), or require the batch-normalization (BN) layers that integer deployment fuses away. We present a forward-only TTA method that operates on deployed, BN-folded, integer-only convolutional networks. The key observation is that fusing BN into the preceding convolution, a mandatory step for integer inference, destroys the statistics that normalization-based adaptation relies on. We restore adaptation by re-normalizing each folded convolution's per-channel output to its clean training statistics, using only forward-pass estimates. The method (i) recovers most of gradient-based TENT's accuracy gain (+20.9 vs. +24.9 points) and matches forward-only BN adaptation, while being the only method that runs on a folded integer-only model; (ii) needs to adapt only 3 of 21 layers (selected without seeing the test corruptions) to recover 93% of the benefit; (iii) survives single-sample streaming with a batch-size-scaled momentum; and (iv) generalizes across three datasets (up to 200 classes) and two architectures. We validate bit-exact int8 convolution execution and deploy on an ESP32-S3, where, measured with a Nordic PPK2 power profiler, the forward-only adaptation (a lightweight fp32 recalibration around the int8 convolutions) costs only 8.3 mJ (6.8% of inference energy) and 21.9 ms on the deployed SIMD-optimized model: forward-only adaptation is cheap on a real microcontroller.

Related papers

• Parameter-Efficient Self-Supervised Adaptation for EEG-FM under Fixed Computational Budgets [0.12891210250935145]
  We investigate whether parameter-efficient self-supervised adaptation, updating only 9% of parameters suffices to align representations to target tasks.<n>We evaluate our method on two state-of-the-art models with different pretraining objectives.<n> SSL adaptation yields consistent gains over linear probing, up to 20x AUCPR.
  arXiv  Detail & Related papers  (2026-08-25T15:38:46Z)
• Speeding Up MACE: Low-Precision Tricks for Equivarient Force Fields [51.95157731126864]
  Machine-learning force fields can deliver accurate molecular dynamics (MD) at high computational cost.<n>This thesis aims to make MACE cheaper and faster by identifying computational bottlenecks and evaluating low-precision execution policies.
  arXiv  Detail & Related papers  (2025-10-23T14:02:34Z)
• POME: Post Optimization Model Edit via Muon-style Projection [74.73326657229347]
  Post-Optimization Model Edit (POME) enhances the performance of fine-tuned large language models.<n>It uses a muon-style projection to $Delta W$, the difference between the fine-tuned and pretrained weights.<n>As a simple post-processing step, POME is completely decoupled from the training pipeline.
  arXiv  Detail & Related papers  (2025-10-08T04:20:11Z)
• MASS: MoErging through Adaptive Subspace Selection [55.03293736484465]
  We present MASS (MoErging through Adaptive Subspace Selection), a new approach to model merging.<n> MASS stores only the most salient singular components for each task and merges them into a shared model.<n>We evaluate MASS on CLIP-based image classification using ViT-B-16, ViT-B-32 and ViT-L-14 for benchmarks of 8, 14 and 20 tasks respectively.
  arXiv  Detail & Related papers  (2025-04-06T08:49:52Z)
• FPN-fusion: Enhanced Linear Complexity Time Series Forecasting Model [1.9780020273082008]
  This study presents a novel time series prediction model, FPN-fusion, designed with linear computational complexity. Our model introduces two key innovations: first, a Feature Pyramid Network (FPN) is employed to effectively capture time series data characteristics, bypassing the traditional decomposition into trend and seasonal components. Empirically, FPN-fusion outperforms DLiner in 31 out of 32 test cases on eight open-source datasets, with an average reduction of 16.8% in mean squared error (MSE) and 11.8% in mean absolute error (MAE)
  arXiv  Detail & Related papers  (2024-06-06T14:34:26Z)
• Test-Time Model Adaptation with Only Forward Passes [68.11784295706995]
  Test-time adaptation has proven effective in adapting a given trained model to unseen test samples with potential distribution shifts. We propose a test-time Forward-Optimization Adaptation (FOA) method. FOA runs on quantized 8-bit ViT, outperforms gradient-based TENT on full-precision 32-bit ViT, and achieves an up to 24-fold memory reduction on ImageNet-C.
  arXiv  Detail & Related papers  (2024-04-02T05:34:33Z)
• Reduced Precision Floating-Point Optimization for Deep Neural Network On-Device Learning on MicroControllers [15.37318446043671]
  This paper introduces a novel reduced precision optimization technique for On-Device Learning (ODL) primitives on MCU-class devices. Our approach results more than two orders of magnitude faster than existing ODL software frameworks for single-core MCUs.
  arXiv  Detail & Related papers  (2023-05-30T16:14:16Z)
• SITA: Single Image Test-time Adaptation [48.789568233682296]
  In Test-time Adaptation (TTA), given a model trained on some source data, the goal is to adapt it to make better predictions for test instances from a different distribution. We consider TTA in a more pragmatic setting which we refer to as SITA (Single Image Test-time Adaptation) Here, when making each prediction, the model has access only to the given single test instance, rather than a batch of instances. We propose a novel approach AugBN for the SITA setting that requires only forward-preserving propagation.
  arXiv  Detail & Related papers  (2021-12-04T15:01:35Z)
• A contextual analysis of multi-layer perceptron models in classifying hand-written digits and letters: limited resources [0.0]
  We extensively test an end-to-end vanilla neural network (MLP) approach in pure numpy without any pre-processing or feature extraction done beforehand. We show that basic data mining operations can significantly improve the performance of the models in terms of computational time.
  arXiv  Detail & Related papers  (2021-07-05T04:30:37Z)
• Tent: Fully Test-time Adaptation by Entropy Minimization [77.85911673550851]
  A model must adapt itself to generalize to new and different data during testing. In this setting of fully test-time adaptation the model has only the test data and its own parameters. We propose to adapt by test entropy minimization (tent): we optimize the model for confidence as measured by the entropy of its predictions.
  arXiv  Detail & Related papers  (2020-06-18T17:55:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.