cs.AINov 17, 2026

BeFit: Shifting Complexity from Weights to Bernstein Activations for Resource-Constrained Sensing

Authors: Malak Gamal El-Din, Yasser Shoukry, Salma Elmalaki

Organizations: Department of Electrical Engineering and Computer Science, University of California, Irvine, CA 92697 USA

Abstract

Time-series sensing increasingly relies on neural networks, yet deployment on resource-constrained microcontrollers remains challenging. Aggressive model compression can reduce accuracy, while restricting weights to multiplication-free representations introduces further losses. We present BeFit, an end-to-end framework that shifts model capacity from weights to learnable Bernstein polynomial activations. BeFit (i) finds compact Bernstein-activated convolutional neural networks (BeCNNs) through hardware-aware neural architecture search and trains them via knowledge distillation; (ii) projects their weights to multiplication-free codes, freezes the codes, and recovers the resulting accuracy loss through the learnable activations; and (iii) generates model-specific kernels that exploit the projected weights for multiplication-free convolutions. On human activity recognition, BeCNNs are 2.75 percentage points (pp) more accurate on average than ReLU CNNs of identical architecture. On a TI CC3220SF (Cortex-M4), weight projection speeds BeCNNs up by 2.2x on average while losing 0.8 pp, compared with 2.4 pp for ReLU CNNs. Against the MLPerf Tiny DS-CNN reference model running on TensorFlow Lite Micro with CMSIS-NN, the unprojected BeCNN achieves 1.0 pp higher accuracy, 2.2x lower latency, and 6.4x lower flash, while the projected model remains within 1.0 pp of the reference accuracy and runs 5.9x faster. In emulation on a RISC-V core without a hardware multiplier, the instruction-count reduction relative to the DS-CNN increases from 5.2x on the Cortex-M4 to 147x. BeFit also extends to transformers and MLPs, and across all tasks the final models stay within 1.56 pp of their float32 teachers on average.

Figures & tables

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Bern2Edge: A Neurosymbolic Compiler for Edge Deployment via Bernstein Polynomial Networks

    Aug 20, 2026Malak Gamal El-Din, Yifan Zhang, Yasser Shoukry +2Edge InferenceFPGA-Based Neural Network Acceleration

  2. Deep Microcompression: Structured Pruning and Bit-packed Quantization for Microcontrollers

    Sep 4, 2026Opegbemi Matthias Busoye, Tolulope Matthew Busoye, Eghonghon-aye EigbeNeural Network CompressionOn-Device Inference

  3. Federated Learning of Binary Neural Networks: Enabling Low-Cost Inference

    Mar 16, 2026Nitin Priyadarshini Shankar, Soham Lahiri, Sheetal Kalyani +1Binary Neural NetworksFederated Learning