Mohammed Alsakabi

Selected Publications

1. arXiv

   JA-SIREN: Deterministic Initialization for Sinusoidal Networks via Spectral Matching

   Mohammed Alsakabi, Kejia Hu, John M. Dolan, and Ozan K. Tonguz

   arXiv preprint arXiv:2606.06671 2026

   Abs arXiv Bib

   Existing implicit neural representation (INR) approaches suffer from stochastic initialization that does not guarantee consistent or high-quality performance across runs, with variations reaching more than 2.5 dB (78%) in image regression. This variation is problematic for scientific computing and simulation, where result reproducibility is crucial. To address this problem, we present Jacobi-Anger Sinusoidal Representation Network (JA-SIREN), a deterministic initialization scheme for sinusoidal networks grounded in classical spectral analysis. By computing the Discrete Sine Transform (DST) of the target signal and leveraging the Jacobi-Anger expansion, we derive closed-form weights for a two-layer sinusoidal MLP that analytically match the network’s initial spectral response to the target signal, requiring no random seed or additional hyperparameter tuning. On the Kodak dataset, JA-SIREN achieves a mean PSNR of 67.18 dB, a 21.30 dB improvement over the best baseline. This is achieved with zero run-to-run variance, confirming that spectrally-informed initialization is a more effective and reproducible alternative to stochastic initialization for sinusoidal INRs.

   @article{alsakabi2026jasiren,
     author = {Alsakabi, Mohammed and Hu, Kejia and Dolan, John M. and Tonguz, Ozan K.},
     title = {JA-SIREN: Deterministic Initialization for Sinusoidal Networks via Spectral Matching},
     journal = {arXiv preprint arXiv:2606.06671},
     year = {2026},
   }

2. arXiv

   FM-SIREN & FM-FINER: Implicit Neural Representation Using Nyquist-based Orthogonality

   Mohammed Alsakabi, Wael Mobeirek, John M. Dolan, and Ozan K. Tonguz

   arXiv preprint arXiv:2509.23438 2025

   Abs arXiv Bib

   Existing periodic activation-based implicit neural representation (INR) networks, such as SIREN and FINER, suffer from hidden feature redundancy, where neurons within a layer capture overlapping frequency components due to the use of a fixed frequency multiplier. This redundancy limits the expressive capacity of multilayer perceptrons (MLPs). Drawing inspiration from classical signal processing methods such as the Discrete Sine Transform (DST), in this paper, we propose FM-SIREN and FM-FINER, which assign Nyquist-informed, neuron-specific frequency multipliers to periodic activations. Contrary to existing approaches, our design introduces frequency diversity without requiring hyperparameter tuning or additional network depth. This simple yet principled approach reduces the redundancy of features by nearly 50% and consistently improves signal reconstruction across diverse INR tasks, such as fitting 1D audio, 2D image and 3D shape, and video, outperforming their baseline counterparts while maintaining efficiency.

   @article{alsakabi2025fm,
     author = {Alsakabi, Mohammed and Mobeirek, Wael and Dolan, John M. and Tonguz, Ozan K.},
     title = {FM-SIREN \& FM-FINER: Implicit Neural Representation Using Nyquist-based Orthogonality},
     journal = {arXiv preprint arXiv:2509.23438},
     year = {2025},
   }

3. ITSC25

   Toward a Low-Cost Perception System in Autonomous Vehicles: A Spectrum Learning Approach

   Mohammed Alsakabi, Aidan Erickson, John M. Dolan, and Ozan K. Tonguz

   In 2025 IEEE Intelligent Transportation Systems Conference (ITSC) 2025

   Abs arXiv Bib

   We present a cost-effective new approach for generating denser depth maps for Autonomous Driving (AD) and Autonomous Vehicles (AVs) by integrating the images obtained from deep neural network (DNN) 4D radar detectors with conventional camera RGB images. Our approach introduces a novel pixel positional encoding algorithm inspired by Bartlett’s spatial spectrum estimation technique. This algorithm transforms both radar depth maps and RGB images into a unified pixel image subspace called the Spatial Spectrum, facilitating effective learning based on their similarities and differences. Our method effectively leverages high-resolution camera images to train radar depth map generative models, addressing the limitations of conventional radar detectors in complex vehicular environments, thus sharpening the radar output. We develop spectrum estimation algorithms tailored for radar depth maps and RGB images, a comprehensive training framework for data-driven generative models, and a camera-radar deployment scheme for AV operation. Our results demonstrate that our approach also outperforms the state-of-the-art (SOTA) by 27.95% in terms of Unidirectional Chamfer Distance (UCD).

   @inproceedings{alsakabi2025toward,
     author = {Alsakabi, Mohammed and Erickson, Aidan and Dolan, John M. and Tonguz, Ozan K.},
     title = {Toward a Low-Cost Perception System in Autonomous Vehicles: A Spectrum Learning Approach},
     booktitle = {2025 IEEE Intelligent Transportation Systems Conference (ITSC)},
     year = {2025},
   }