Shaobo Han 2023 NEC Labs America

Shaobo Han

Senior Researcher

Optical Networking & Sensing

Posts

Exploring Compositional Visual Generation with Latent Classifier Guidance

Exploring Compositional Visual Generation with Latent Classifier Guidance Diffusion probabilistic models have achieved enormous success in the field of image generation and manipulation. In this paper, we explore a novel paradigm of using the diffusion model and classifier guidance in the latent semantic space for compositional visual tasks. Specifically, we train latent diffusion models and auxiliary latent classifiers to facilitate non-linear navigation of latent representation generation for any pre-trained generative model with a semantic latent space. We demonstrate that such conditional generation achieved by latent classifier guidance provably maximizes a lower bound of the conditional log probability during training. To maintain the original semantics during manipulation, we introduce a new guidance term, which we show is crucial for achieving compositionality. With additional assumptions, we show that the non-linear manipulation reduces to a simple latent arithmetic approach. We show that this paradigm based on latent classifier guidance is agnostic to pre-trained generative models, and present competitive results for both image generation and sequential manipulation of real and synthetic images. Our findings suggest that latent classifier guidance is a promising approach that merits further exploration, even in the presence of other strong competing methods.

Explore Benefits of Distributed Fiber Optic Sensing for Optical Network Service Providers

Explore Benefits of Distributed Fiber Optic Sensing for Optical Network Service Providers We review various applications of distributed fiber optic sensing (DFOS) and machine learning (ML) technologies that particularly benefit telecom operators’ fiber networks and businesses. By leveraging relative phase shift of the reflectance of coherent Rayleigh, Brillouin and Raman scattering of light wave, the ambient environmental vibration, acoustic effects, temperature and fiber/cable strain can be detected. Fiber optic sensing technology allows optical fiber to support sensing features in addition to its conventional role to transmit data in telecommunications. DFOS has recently helped telecom operators by adding multiple sensing features and proved feasibility of co-existence of sensing and communication systems on same fiber. We review the architecture of DFOS technique and show examples where optical fiber sensing helps enhance network operation efficiency and create new services for customers on deployed fiber infrastructures, such as determination of cable locations, cable cut prevention, perimeter intrusion detection and networked sensing applications. In addition, edge AI platform allows data processing to be conducted on-the-fly with low latency. Based on discriminative spatial-temporal signatures of different events of interest, real-time processing of the sensing data from the DFOS system provides results of the detection, classification and localization immediately.

Improvement of resilience of submarine networks based on fiber sensing

Improvement of resilience of submarine networks based on fiber sensing Simultaneous phase and polarization sensing with span length resolution using the supervisory path is demonstrated. It is shown that by measuring polarization rotation matrix of the return paths, instead of monitoring only the state of polarization, location of the polarization disturbance can be determined even for large polarization rotations. By using the polarization rotation matrices, the phase and polarization disturbances are successfully decoupled. How the existing supervisory system and sensing can coexist in new SDM cables that utilizes pump sharing is discussed.

Polarization Sensing Using Polarization Rotation Matrix Eigenvalue Method

Polarization Sensing Using Polarization Rotation Matrix Eigenvalue Method Polarization-based, multi-span sensing over a link with reflection-back circuits is demonstrated experimentally. By measuring rotation matrices instead of just monitoring polarization, a 35 dB extinction in localization is achieved regardless of the disturbance magnitude.

Ambient Noise based Weakly Supervised Manhole Localization Methods over Deployed Fiber Networks

Ambient Noise based Weakly Supervised Manhole Localization Methods over Deployed Fiber Networks We present a manhole localization method based on distributed fiber optic sensing and weakly supervised machine learning techniques. For the first time to our knowledge, ambient environment data is used for underground cable mapping with the promise of enhancing operational efficiency and reducing field work. To effectively accommodate the weak informativeness of ambient data, a selective data sampling scheme and an attention-based deep multiple instance classification model are adopted, which only requires weakly annotated data. The proposed approach is validated on field data collected by a fiber sensing system over multiple existing fiber networks.

Using Global Fiber Networks for Environmental Sensing

Using Global Fiber Networks for Environmental Sensing We review recent advances in distributed fiber optic sensing (DFOS) and their applications. The scattering mechanisms in glass, which are exploited for reflectometry-based DFOS, are Rayleigh, Brillouin, and Raman scatterings. These are sensitive to either strain and/or temperature, allowing optical fiber cables to monitor their ambient environment in addition to their conventional role as a medium for telecommunications. Recently, DFOS leveraged technologies developed for telecommunications, such as coherent detection, digital signal processing, coding, and spatial/frequency diversity, to achieve improved performance in terms of measurand resolution, reach, spatial resolution, and bandwidth. We review the theory and architecture of commonly used DFOS methods. We provide recent experimental and field trial results where DFOS was used in wide-ranging applications, such as geohazard monitoring, seismic monitoring, traffic monitoring, and infrastructure health monitoring. Events of interest often have unique signatures either in the spatial, temporal, frequency, or wavenumber domains. Based on the temperature and strain raw data obtained from DFOS, downstream postprocessing allows the detection, classification, and localization of events. Combining DFOS with machine learning methods, it is possible to realize complete sensor systems that are compact, low cost, and can operate in harsh environments and difficult-to-access locations, facilitating increased public safety and smarter cities.

Field Trials of Vibration Detection, Localization and Classification over Deployed Telecom Fiber Cables

Field Trials of Vibration Detection, Localization and Classification over Deployed Telecom Fiber Cables We review sensing fusion results of integrating fiber sensing with video for machine-learning-based localization and classification of impulsive acoustic event detection. Classification accuracy >97% was achieved on aerial coils, and >99% using fiber-based signal enhancers.

Evolution of Fiber Infrastructure – From Data Transmission to Network Sensing

Evolution of Fiber Infrastructure – From Data Transmission to Network Sensing We review multiple use cases over deployed networks including co-existing sensing/data transmission, cable cut prevention and perimeter intrusion detection to realize telecom infrastructure can be sensing backbones instead of the sole function of data transmission.

Field Tests of Impulsive Acoustic Event Detection, Localization, and Classification Over Telecom Fiber Networks

Field Tests of Impulsive Acoustic Event Detection, Localization, and Classification Over Telecom Fiber Networks We report distributed-fiber-optic-sensing results on impulsive acoustic events localization/classification over telecom networks. A deep-learning-based model was trained to classify starter-gun and fireworks signatures with high accuracy of > 99% using fiber-based-signal-enhancer and >97% using aerial coils.

Provable Adaptation Across Multiway Domains via Representation Learning

Provable Adaptation Across Multiway Domains via Representation Learning This paper studies zero-shot domain adaptation where each domain is indexed on a multi-dimensional array, and we only have data from a small subset of domains. Our goal is to produce predictors that perform well on unseen domains. We propose a model which consists of a domain-invariant latent representation layer and a domain-specific linear prediction layer with a low-rank tensor structure. Theoretically, we present explicit sample complexity bounds to characterize the prediction error on unseen domains in terms of the number of domains with training data and the number of data per domain. To our knowledge, this is the first finite-sample guarantee for zero-shot domain adaptation. In addition, we provide experiments on two-way MNIST and four-way fiber sensing datasets to demonstrate the effectiveness of our proposed model.