arXiv subjects

Elad Cohen

Publications and source records attributed to Elad Cohen.

7 recordsLinked to original sources

arXiv2026

Low-Latency State Space Voice Activity Detection with Robust Onset Time Evaluation

Voice Activity Detection (VAD) systems are commonly evaluated using metrics such as the area under the receiver operating characteristic curve (AUROC), but these metrics do not account for temporal responsiveness. For low-latency applications, however, accurately measuring speech onset delay is essential. This is particularly challenging because onset latency evaluation is affected by noise and systematic misalignment in annotation timestamps. In this work, we introduce a probabilistic framework for evaluating VAD onset time under noisy temporal labels. We model annotated onset times as noisy observations of latent acoustic onsets and estimate the resulting discrepancy distribution from data. We show that this approach provides a more stable and robust estimate of algorithmic latency. In addition, we introduce S4VAD, the first state-space-model-based (SSM-based) VAD architecture. S4VAD supports efficient streaming inference while directly controlling the decay of past acoustic evidence, enabling fast and responsive speech-onset detection. We evaluate VAD onset latency across several low-latency architectures, including CNN, Transformer, and RNN models. Our proposed VAD achieves the lowest latency while maintaining a competitive AUROC.

eess.AS↗

arXiv2025

MLoRQ: Bridging Low-Rank and Quantization for Transformer Compression

Deploying transformer-based neural networks on resource-constrained edge devices presents a significant challenge. This challenge is often addressed through various techniques, such as low-rank approximation and mixed-precision quantization. In this work, we introduce Mixed Low-Rank and Quantization (MLoRQ), a novel method that integrates both techniques. MLoRQ employs a two-stage optimization process to determine optimal bit-width and rank assignments for each layer, adhering to predefined memory constraints. This process includes: (i) an intra-layer optimization that identifies potentially optimal compression solutions out of all low-rank and quantization combinations; (ii) an inter-layer optimization that assigns bit-width precision and rank to each layer while ensuring the memory constraint is met. An optional final step applies a sequential optimization process using a modified adaptive rounding technique to mitigate compression-induced errors in joint low-rank approximation and quantization. The method is compatible and can be seamlessly integrated with most existing quantization algorithms. MLoRQ shows state-of-the-art results with up to 15\% performance improvement, evaluated on Vision Transformers for image classification, object detection, and instance segmentation tasks.

cs.LG↗

arXiv2025

Efficient Image Restoration via Latent Consistency Flow Matching

Recent advances in generative image restoration (IR) have demonstrated impressive results. However, these methods are hindered by their substantial size and computational demands, rendering them unsuitable for deployment on edge devices. This work introduces ELIR, an Efficient Latent Image Restoration method. ELIR addresses the distortion-perception trade-off within the latent space and produces high-quality images using a latent consistency flow-based model. In addition, ELIR introduces an efficient and lightweight architecture. Consequently, ELIR is 4$\times$ smaller and faster than state-of-the-art diffusion and flow-based approaches for blind face restoration, enabling a deployment on resource-constrained devices. Comprehensive evaluations of various image restoration tasks and datasets show that ELIR achieves competitive performance compared to state-of-the-art methods, effectively balancing distortion and perceptual quality metrics while significantly reducing model size and computational cost. The code is available at: https://github.com/eladc-git/ELIR

eess.IV↗

arXiv2023

EPTQ: Enhanced Post-Training Quantization via Hessian-guided Network-wise Optimization

Quantization is a key method for deploying deep neural networks on edge devices with limited memory and computation resources. Recent improvements in Post-Training Quantization (PTQ) methods were achieved by an additional local optimization process for learning the weight quantization rounding policy. However, a gap exists when employing network-wise optimization with small representative datasets. In this paper, we propose a new method for enhanced PTQ (EPTQ) that employs a network-wise quantization optimization process, which benefits from considering cross-layer dependencies during optimization. EPTQ enables network-wise optimization with a small representative dataset using a novel sample-layer attention score based on a label-free Hessian matrix upper bound. The label-free approach makes our method suitable for the PTQ scheme. We give a theoretical analysis for the said bound and use it to construct a knowledge distillation loss that guides the optimization to focus on the more sensitive layers and samples. In addition, we leverage the Hessian upper bound to improve the weight quantization parameters selection by focusing on the more sensitive elements in the weight tensors. Empirically, by employing EPTQ we achieve state-of-the-art results on various models, tasks, and datasets, including ImageNet classification, COCO object detection, and Pascal-VOC for semantic segmentation.

cs.CV↗

arXiv2021

HPTQ: Hardware-Friendly Post Training Quantization

Neural network quantization enables the deployment of models on edge devices. An essential requirement for their hardware efficiency is that the quantizers are hardware-friendly: uniform, symmetric, and with power-of-two thresholds. To the best of our knowledge, current post-training quantization methods do not support all of these constraints simultaneously. In this work, we introduce a hardware-friendly post training quantization (HPTQ) framework, which addresses this problem by synergistically combining several known quantization methods. We perform a large-scale study on four tasks: classification, object detection, semantic segmentation and pose estimation over a wide variety of network architectures. Our extensive experiments show that competitive results can be obtained under hardware-friendly constraints.

cs.CV↗

arXiv2012

Phase transition in crowd synchrony of delay-coupled multilayer laser networks

An analogy between crowd synchrony and multi-layer neural network architectures is proposed. It indicates that many non-identical dynamical elements (oscillators) communicating indirectly via a few mediators (hubs) can synchronize when the number of delayed couplings to the hubs or the strength of the couplings is large enough. This phenomenon is modeled using a system of semiconductor lasers optically delay-coupled in either a fully connected or a diluted manner to a fixed number of non-identical central hub lasers. A universal phase transition to crowd synchrony with hysteresis is observed, where the time to achieve synchronization diverges near the critical coupling independent of the number of hubs.

nlin.CD↗

arXiv2007

Spiking Optical Patterns and Synchronization

We analyze the time resolved spike statistics of a solitary and two mutually interacting chaotic semiconductor lasers whose chaos is characterized by apparently random, short intensity spikes. Repulsion between two successive spikes is observed, resulting in a refractory period which is largest at laser threshold. For time intervals between spikes greater than the refractory period, the distribution of the intervals follows a Poisson distribution. The spiking pattern is highly periodic over time windows corresponding to the optical length of the external cavity, with a slow change of the spiking pattern as time increases. When zero-lag synchronization between the two lasers is established, the statistics of the nearly perfectly matched spikes are not altered. The similarity of these features to those found in complex interacting neural networks, suggests the use of laser systems as simpler physical models for neural networks.

physics.optics↗