-
SpecMoE: Spectral Mixture-of-Experts Foundation Model for Cross-Species EEG Decoding
Authors:
Davy Darankoum,
Chloé Habermacher,
Julien Volle,
Sergei Grudinin
Abstract:
Decoding the orchestration of neural activity in electroencephalography (EEG) signals is a central challenge in bridging neuroscience with artificial intelligence. Foundation models have made strides in generalized EEG decoding, yet many existing frameworks primarily relying on separate temporal and spectral masking of raw signals during self-supervised pretraining. Such strategies often tend to b…
▽ More
Decoding the orchestration of neural activity in electroencephalography (EEG) signals is a central challenge in bridging neuroscience with artificial intelligence. Foundation models have made strides in generalized EEG decoding, yet many existing frameworks primarily relying on separate temporal and spectral masking of raw signals during self-supervised pretraining. Such strategies often tend to bias learning toward high-frequency oscillations, as low-frequency rhythmic patterns can be easily inferred from the unmasked signal.
We introduce a foundation model that utilizes a novel Gaussian-smoothed masking scheme applied to short-time Fourier transform (STFT) maps. By jointly applying time, frequency, and time-frequency Gaussian masks, we make the reconstruction task much more challenging, forcing the model to learn intricate neural patterns across both high- and low-frequency domains. To effectively recover signals under this aggressive masking strategy, we design SpecHi-Net, a U-shaped hierarchical architecture with multiple encoding and decoding stages. To accelerate large-scale pretraining, we partition the data into three subsets, each used to train an independent expert model. We then combine these models through SpecMoE, a mixture of experts framework guided by a learned spectral gating mechanism.
SpecMoE achieves state-of-the-art performance across a diverse set of EEG decoding tasks, including sleep staging, emotion recognition, motor imagery classification, abnormal signal detection, and drug effect prediction. Importantly, the model demonstrates strong cross-species and cross-subject generalization, maintaining high accuracy on both human and murine EEG datasets.
△ Less
Submitted 30 March, 2026; v1 submitted 17 March, 2026;
originally announced March 2026.
-
CoSupFormer : A Contrastive Supervised learning approach for EEG signal Classification
Authors:
D. Darankoum,
C. Habermacher,
J. Volle,
S. Grudinin
Abstract:
Electroencephalography signals (EEGs) contain rich multi-scale information crucial for understanding brain states, with potential applications in diagnosing and advancing the drug development landscape. However, extracting meaningful features from raw EEG signals while handling noise and channel variability remains a major challenge. This work proposes a novel end-to-end deep-learning framework th…
▽ More
Electroencephalography signals (EEGs) contain rich multi-scale information crucial for understanding brain states, with potential applications in diagnosing and advancing the drug development landscape. However, extracting meaningful features from raw EEG signals while handling noise and channel variability remains a major challenge. This work proposes a novel end-to-end deep-learning framework that addresses these issues through several key innovations. First, we designed an encoder capable of explicitly capturing multi-scale frequency oscillations covering a wide range of features for different EEG-related tasks. Secondly, to model complex dependencies and handle the high temporal resolution of EEGs, we introduced an attention-based encoder that simultaneously learns interactions across EEG channels and within localized {\em patches} of individual channels. We integrated a dedicated gating network on top of the attention encoder to dynamically filter out noisy and non-informative channels, enhancing the reliability of EEG data. The entire encoding process is guided by a novel loss function, which leverages supervised and contrastive learning, significantly improving model generalization. We validated our approach in multiple applications, ranging from the classification of effects across multiple Central Nervous System (CNS) disorders treatments to the diagnosis of Parkinson's and Alzheimer's disease. Our results demonstrate that the proposed learning paradigm can extract biologically meaningful patterns from raw EEG signals across different species, autonomously select high-quality channels, and achieve robust generalization through innovative architectural and loss design.
△ Less
Submitted 24 September, 2025;
originally announced September 2025.
-
Challenges in Non-Polymeric Crystal Structure Prediction: Why a Geometric, Permutation-Invariant Loss is Needed
Authors:
Emmanuel Jehanno,
Romain Menegaux,
Julien Mairal,
Sergei Grudinin
Abstract:
Crystalline structure prediction is an essential prerequisite for designing materials with targeted properties. Yet, it is still an open challenge in materials design and drug discovery. Despite recent advances in computational materials science, accurately predicting three-dimensional non-polymeric crystal structures remains elusive. In this work, we focus on the molecular assembly problem, where…
▽ More
Crystalline structure prediction is an essential prerequisite for designing materials with targeted properties. Yet, it is still an open challenge in materials design and drug discovery. Despite recent advances in computational materials science, accurately predicting three-dimensional non-polymeric crystal structures remains elusive. In this work, we focus on the molecular assembly problem, where a set $\mathcal{S}$ of identical rigid molecules is packed to form a crystalline structure. Such a simplified formulation provides a useful approximation to the actual problem. However, while recent state-of-the-art methods have increasingly adopted sophisticated techniques, the underlying learning objective remains ill-posed. We propose a better formulation that introduces a loss function capturing key geometric molecular properties while ensuring permutation invariance over $\mathcal{S}$. Remarkably, we demonstrate that within this framework, a simple regression model already outperforms prior approaches, including flow matching techniques, on the COD-Cluster17 benchmark, a curated non-polymeric subset of the Crystallography Open Database (COD).
△ Less
Submitted 26 September, 2025; v1 submitted 31 August, 2025;
originally announced September 2025.
-
PETIMOT: A Novel Framework for Inferring Protein Motions from Sparse Data Using SE(3)-Equivariant Graph Neural Networks
Authors:
Valentin Lombard,
Sergei Grudinin,
Elodie Laine
Abstract:
Proteins move and deform to ensure their biological functions. Despite significant progress in protein structure prediction, approximating conformational ensembles at physiological conditions remains a fundamental open problem. This paper presents a novel perspective on the problem by directly targeting continuous compact representations of protein motions inferred from sparse experimental observa…
▽ More
Proteins move and deform to ensure their biological functions. Despite significant progress in protein structure prediction, approximating conformational ensembles at physiological conditions remains a fundamental open problem. This paper presents a novel perspective on the problem by directly targeting continuous compact representations of protein motions inferred from sparse experimental observations. We develop a task-specific loss function enforcing data symmetries, including scaling and permutation operations. Our method PETIMOT (Protein sEquence and sTructure-based Inference of MOTions) leverages transfer learning from pre-trained protein language models through an SE(3)-equivariant graph neural network. When trained and evaluated on the Protein Data Bank, PETIMOT shows superior performance in time and accuracy, capturing protein dynamics, particularly large/slow conformational changes, compared to state-of-the-art flow-matching approaches and traditional physics-based models.
△ Less
Submitted 19 March, 2025;
originally announced April 2025.
-
BAnG: Bidirectional Anchored Generation for Conditional RNA Design
Authors:
Roman Klypa,
Alberto Bietti,
Sergei Grudinin
Abstract:
Designing RNA molecules that interact with specific proteins is a critical challenge in experimental and computational biology. Existing computational approaches require a substantial amount of previously known interacting RNA sequences for each specific protein or a detailed knowledge of RNA structure, restricting their utility in practice. To address this limitation, we develop RNA-BAnG, a deep…
▽ More
Designing RNA molecules that interact with specific proteins is a critical challenge in experimental and computational biology. Existing computational approaches require a substantial amount of previously known interacting RNA sequences for each specific protein or a detailed knowledge of RNA structure, restricting their utility in practice. To address this limitation, we develop RNA-BAnG, a deep learning-based model designed to generate RNA sequences for protein interactions without these requirements. Central to our approach is a novel generative method, Bidirectional Anchored Generation (BAnG), which leverages the observation that protein-binding RNA sequences often contain functional binding motifs embedded within broader sequence contexts. We first validate our method on generic synthetic tasks involving similar localized motifs to those appearing in RNAs, demonstrating its benefits over existing generative approaches. We then evaluate our model on biological sequences, showing its effectiveness for conditional RNA sequence design given a binding protein.
△ Less
Submitted 23 June, 2025; v1 submitted 28 February, 2025;
originally announced February 2025.
-
From Epilepsy Seizures Classification to Detection: A Deep Learning-based Approach for Raw EEG Signals
Authors:
Davy Darankoum,
Manon Villalba,
Clelia Allioux,
Baptiste Caraballo,
Carine Dumont,
Eloise Gronlier,
Corinne Roucard,
Yann Roche,
Chloe Habermacher,
Sergei Grudinin,
Julien Volle
Abstract:
Epilepsy represents the most prevalent neurological disease in the world. One-third of people suffering from mesial temporal lobe epilepsy (MTLE) exhibit drug resistance, urging the need to develop new treatments. A key part in anti-seizure medication (ASM) development is the capability of detecting and quantifying epileptic seizures occurring in electroencephalogram (EEG) signals, which is crucia…
▽ More
Epilepsy represents the most prevalent neurological disease in the world. One-third of people suffering from mesial temporal lobe epilepsy (MTLE) exhibit drug resistance, urging the need to develop new treatments. A key part in anti-seizure medication (ASM) development is the capability of detecting and quantifying epileptic seizures occurring in electroencephalogram (EEG) signals, which is crucial for treatment efficacy evaluation. In this study, we introduced a seizure detection pipeline based on deep learning models applied to raw EEG signals. This pipeline integrates: a new pre-processing technique which segments continuous raw EEG signals without prior distinction between seizure and seizure-free activities; a post-processing algorithm developed to reassemble EEG segments and allow the identification of seizures start/end; and finally, a new evaluation procedure based on a strict seizure events comparison between predicted and real labels. Models training have been performed using a data splitting strategy which addresses the potential for data leakage. We demonstrated the fundamental differences between a seizure classification and a seizure detection task and showed the differences in performance between the two tasks. Finally, we demonstrated the generalization capabilities across species of our best architecture, combining a Convolutional Neural Network and a Transformer encoder. The model was trained on animal EEGs and tested on human EEGs with a F1-score of 93% on a balanced Bonn dataset.
△ Less
Submitted 2 November, 2025; v1 submitted 4 October, 2024;
originally announced October 2024.
-
On the Fourier analysis in the SO(3) space : EquiLoPO Network
Authors:
Dmitrii Zhemchuzhnikov,
Sergei Grudinin
Abstract:
Analyzing volumetric data with rotational invariance or equivariance is an active topic in current research. Existing deep-learning approaches utilize either group convolutional networks limited to discrete rotations or steerable convolutional networks with constrained filter structures. This work proposes a novel equivariant neural network architecture that achieves analytical Equivariance to Loc…
▽ More
Analyzing volumetric data with rotational invariance or equivariance is an active topic in current research. Existing deep-learning approaches utilize either group convolutional networks limited to discrete rotations or steerable convolutional networks with constrained filter structures. This work proposes a novel equivariant neural network architecture that achieves analytical Equivariance to Local Pattern Orientation on the continuous SO(3) group while allowing unconstrained trainable filters - EquiLoPO Network. Our key innovations are a group convolutional operation leveraging irreducible representations as the Fourier basis and a local activation function in the SO(3) space that provides a well-defined mapping from input to output functions, preserving equivariance. By integrating these operations into a ResNet-style architecture, we propose a model that overcomes the limitations of prior methods. A comprehensive evaluation on diverse 3D medical imaging datasets from MedMNIST3D demonstrates the effectiveness of our approach, which consistently outperforms state of the art. This work suggests the benefits of true rotational equivariance on SO(3) and flexible unconstrained filters enabled by the local activation function, providing a flexible framework for equivariant deep learning on volumetric data with potential applications across domains. Our code is publicly available at https://gricad-gitlab.univ-grenoble-alpes.fr/GruLab/ILPO/-/tree/main/EquiLoPO.
△ Less
Submitted 19 March, 2025; v1 submitted 24 April, 2024;
originally announced April 2024.
-
ILPO-NET: Network for the invariant recognition of arbitrary volumetric patterns in 3D
Authors:
Dmitrii Zhemchuzhnikov,
Sergei Grudinin
Abstract:
Effective recognition of spatial patterns and learning their hierarchy is crucial in modern spatial data analysis. Volumetric data applications seek techniques ensuring invariance not only to shifts but also to pattern rotations. While traditional methods can readily achieve translational invariance, rotational invariance possesses multiple challenges and remains an active area of research. Here,…
▽ More
Effective recognition of spatial patterns and learning their hierarchy is crucial in modern spatial data analysis. Volumetric data applications seek techniques ensuring invariance not only to shifts but also to pattern rotations. While traditional methods can readily achieve translational invariance, rotational invariance possesses multiple challenges and remains an active area of research. Here, we present ILPO-Net (Invariant to Local Patterns Orientation Network), a novel approach that handles arbitrarily shaped patterns with the convolutional operation inherently invariant to local spatial pattern orientations using the Wigner matrix expansions. Our architecture seamlessly integrates the new convolution operator and, when benchmarked on diverse volumetric datasets such as MedMNIST and CATH, demonstrates superior performance over the baselines with significantly reduced parameter counts - up to 1000 times fewer in the case of MedMNIST. Beyond these demonstrations, ILPO-Net's rotational invariance paves the way for other applications across multiple disciplines. Our code is publicly available at https://gricad-gitlab.univ-grenoble-alpes.fr/GruLab/ILPO/-/tree/main/ILPONet.
△ Less
Submitted 24 April, 2024; v1 submitted 28 March, 2024;
originally announced March 2024.
-
6DCNN with roto-translational convolution filters for volumetric data processing
Authors:
Dmitrii Zhemchuzhnikov,
Ilia Igashov,
Sergei Grudinin
Abstract:
In this work, we introduce 6D Convolutional Neural Network (6DCNN) designed to tackle the problem of detecting relative positions and orientations of local patterns when processing three-dimensional volumetric data. 6DCNN also includes SE(3)-equivariant message-passing and nonlinear activation operations constructed in the Fourier space. Working in the Fourier space allows significantly reducing t…
▽ More
In this work, we introduce 6D Convolutional Neural Network (6DCNN) designed to tackle the problem of detecting relative positions and orientations of local patterns when processing three-dimensional volumetric data. 6DCNN also includes SE(3)-equivariant message-passing and nonlinear activation operations constructed in the Fourier space. Working in the Fourier space allows significantly reducing the computational complexity of our operations. We demonstrate the properties of the 6D convolution and its efficiency in the recognition of spatial patterns. We also assess the 6DCNN model on several datasets from the recent CASP protein structure prediction challenges. Here, 6DCNN improves over the baseline architecture and also outperforms the state of the art.
△ Less
Submitted 30 July, 2021; v1 submitted 26 July, 2021;
originally announced July 2021.
-
Protein sequence-to-structure learning: Is this the end(-to-end revolution)?
Authors:
Elodie Laine,
Stephan Eismann,
Arne Elofsson,
Sergei Grudinin
Abstract:
The potential of deep learning has been recognized in the protein structure prediction community for some time, and became indisputable after CASP13. In CASP14, deep learning has boosted the field to unanticipated levels reaching near-experimental accuracy. This success comes from advances transferred from other machine learning areas, as well as methods specifically designed to deal with protein…
▽ More
The potential of deep learning has been recognized in the protein structure prediction community for some time, and became indisputable after CASP13. In CASP14, deep learning has boosted the field to unanticipated levels reaching near-experimental accuracy. This success comes from advances transferred from other machine learning areas, as well as methods specifically designed to deal with protein sequences and structures, and their abstractions. Novel emerging approaches include (i) geometric learning, i.e. learning on representations such as graphs, 3D Voronoi tessellations, and point clouds; (ii) pre-trained protein language models leveraging attention; (iii) equivariant architectures preserving the symmetry of 3D space; (iv) use of large meta-genome databases; (v) combinations of protein representations; (vi) and finally truly end-to-end architectures, i.e. differentiable models starting from a sequence and returning a 3D structure. Here, we provide an overview and our opinion of the novel deep learning approaches developed in the last two years and widely used in CASP14.
△ Less
Submitted 13 September, 2021; v1 submitted 16 May, 2021;
originally announced May 2021.
-
Spherical convolutions on molecular graphs for protein model quality assessment
Authors:
Ilia Igashov,
Nikita Pavlichenko,
Sergei Grudinin
Abstract:
Processing information on 3D objects requires methods stable to rigid-body transformations, in particular rotations, of the input data. In image processing tasks, convolutional neural networks achieve this property using rotation-equivariant operations. However, contrary to images, graphs generally have irregular topology. This makes it challenging to define a rotation-equivariant convolution oper…
▽ More
Processing information on 3D objects requires methods stable to rigid-body transformations, in particular rotations, of the input data. In image processing tasks, convolutional neural networks achieve this property using rotation-equivariant operations. However, contrary to images, graphs generally have irregular topology. This makes it challenging to define a rotation-equivariant convolution operation on these structures. In this work, we propose Spherical Graph Convolutional Network (S-GCN) that processes 3D models of proteins represented as molecular graphs. In a protein molecule, individual amino acids have common topological elements. This allows us to unambiguously associate each amino acid with a local coordinate system and construct rotation-equivariant spherical filters that operate on angular information between graph nodes. Within the framework of the protein model quality assessment problem, we demonstrate that the proposed spherical convolution method significantly improves the quality of model assessment compared to the standard message-passing approach. It is also comparable to state-of-the-art methods, as we demonstrate on Critical Assessment of Structure Prediction (CASP) benchmarks. The proposed technique operates only on geometric features of protein 3D models. This makes it universal and applicable to any other geometric-learning task where the graph structure allows constructing local coordinate systems.
△ Less
Submitted 6 January, 2021; v1 submitted 16 November, 2020;
originally announced November 2020.
-
DeepSymmetry : Using 3D convolutional networks for identification of tandem repeats and internal symmetries in protein structures
Authors:
Guillaume Pagès,
Sergei Grudinin
Abstract:
Motivation: Thanks to the recent advances in structural biology, nowadays three-dimensional structures of various proteins are solved on a routine basis. A large portion of these contain structural repetitions or internal symmetries. To understand the evolution mechanisms of these proteins and how structural repetitions affect the protein function, we need to be able to detect such proteins very r…
▽ More
Motivation: Thanks to the recent advances in structural biology, nowadays three-dimensional structures of various proteins are solved on a routine basis. A large portion of these contain structural repetitions or internal symmetries. To understand the evolution mechanisms of these proteins and how structural repetitions affect the protein function, we need to be able to detect such proteins very robustly. As deep learning is particularly suited to deal with spatially organized data, we applied it to the detection of proteins with structural repetitions. Results: We present DeepSymmetry, a versatile method based on three-dimensional (3D) convolutional networks that detects structural repetitions in proteins and their density maps. Our method is designed to identify tandem repeat proteins, proteins with internal symmetries, symmetries in the raw density maps, their symmetry order, and also the corresponding symmetry axes. Detection of symmetry axes is based on learning six-dimensional Veronese mappings of 3D vectors, and the median angular error of axis determination is less than one degree. We demonstrate the capabilities of our method on benchmarks with tandem repeated proteins and also with symmetrical assemblies. For example, we have discovered over 10,000 putative tandem repeat proteins that are not currently present in the RepeatsDB database. Availability: The method is available at https://team.inria.fr/nano-d/software/deepsymmetry. It consists of a C++ executable that transforms molecular structures into volumetric density maps, and a Python code based on the TensorFlow framework for applying the DeepSymmetry model to these maps.
△ Less
Submitted 29 October, 2018;
originally announced October 2018.
-
Deep convolutional networks for quality assessment of protein folds
Authors:
Georgy Derevyanko,
Sergei Grudinin,
Yoshua Bengio,
Guillaume Lamoureux
Abstract:
The computational prediction of a protein structure from its sequence generally relies on a method to assess the quality of protein models. Most assessment methods rank candidate models using heavily engineered structural features, defined as complex functions of the atomic coordinates. However, very few methods have attempted to learn these features directly from the data. We show that deep convo…
▽ More
The computational prediction of a protein structure from its sequence generally relies on a method to assess the quality of protein models. Most assessment methods rank candidate models using heavily engineered structural features, defined as complex functions of the atomic coordinates. However, very few methods have attempted to learn these features directly from the data. We show that deep convolutional networks can be used to predict the ranking of model structures solely on the basis of their raw three-dimensional atomic densities, without any feature tuning. We develop a deep neural network that performs on par with state-of-the-art algorithms from the literature. The network is trained on decoys from the CASP7 to CASP10 datasets and its performance is tested on the CASP11 dataset. On the CASP11 stage 2 dataset, it achieves a loss of 0.064, whereas the best performing method achieves a loss of 0.063. Additional testing on decoys from the CASP12, CAMEO, and 3DRobot datasets confirms that the network performs consistently well across a variety of protein structures. While the network learns to assess structural decoys globally and does not rely on any predefined features, it can be analyzed to show that it implicitly identifies regions that deviate from the native structure.
△ Less
Submitted 18 January, 2018;
originally announced January 2018.
-
Searching for Exoplanets Using a Microresonator Astrocomb
Authors:
Myoung-Gyun Suh,
Xu Yi,
Yu-Hung Lai,
S. Leifer,
Ivan S. Grudinin,
G. Vasisht,
Emily C. Martin,
Michael P. Fitzgerald,
G. Doppmann,
J. Wang,
D. Mawet,
Scott B. Papp,
Scott A. Diddams,
C. Beichman,
Kerry Vahala
Abstract:
Detection of weak radial velocity shifts of host stars induced by orbiting planets is an important technique for discovering and characterizing planets beyond our solar system. Optical frequency combs enable calibration of stellar radial velocity shifts at levels required for detection of Earth analogs. A new chip-based device, the Kerr soliton microcomb, has properties ideal for ubiquitous applic…
▽ More
Detection of weak radial velocity shifts of host stars induced by orbiting planets is an important technique for discovering and characterizing planets beyond our solar system. Optical frequency combs enable calibration of stellar radial velocity shifts at levels required for detection of Earth analogs. A new chip-based device, the Kerr soliton microcomb, has properties ideal for ubiquitous application outside the lab and even in future space-borne instruments. Moreover, microcomb spectra are ideally suited for astronomical spectrograph calibration and eliminate filtering steps required by conventional mode-locked-laser frequency combs. Here, for the calibration of astronomical spectrographs, we demonstrate an atomic/molecular line-referenced, near-infrared soliton microcomb. Efforts to search for the known exoplanet HD 187123b were conducted at the Keck-II telescope as a first in-the-field demonstration of microcombs.
△ Less
Submitted 16 January, 2018;
originally announced January 2018.
-
Quadratic Programming Approach to Fit Protein Complexes into Electron Density Maps
Authors:
Roman Pogodin,
Alexander Katrutsa,
Sergei Grudinin
Abstract:
The paper investigates the problem of fitting protein complexes into electron density maps. They are represented by high-resolution cryoEM density maps converted into overlapping matrices and partly show a structure of a complex. The general purpose is to define positions of all proteins inside it. This problem is known to be NP-hard, since it lays in the field of combinatorial optimization over a…
▽ More
The paper investigates the problem of fitting protein complexes into electron density maps. They are represented by high-resolution cryoEM density maps converted into overlapping matrices and partly show a structure of a complex. The general purpose is to define positions of all proteins inside it. This problem is known to be NP-hard, since it lays in the field of combinatorial optimization over a set of discrete states of the complex. We introduce quadratic programming approaches to the problem. To find an approximate solution, we convert a density map into an overlapping matrix, which is generally indefinite. Since the matrix is indefinite, the optimization problem for the corresponding quadratic form is non-convex. To treat non-convexity of the optimization problem, we use different convex relaxations to find which set of proteins minimizes the quadratic form best.
△ Less
Submitted 9 January, 2017;
originally announced January 2017.
-
Inverse Protein Folding Problem via Quadratic Programming
Authors:
Andrii Riazanov,
Mikhail Karasikov,
Sergei Grudinin
Abstract:
This paper presents a method of reconstruction a primary structure of a protein that folds into a given geometrical shape. This method predicts the primary structure of a protein and restores its linear sequence of amino acids in the polypeptide chain using the tertiary structure of a molecule. Unknown amino acids are determined according to the principle of energy minimization. This study represe…
▽ More
This paper presents a method of reconstruction a primary structure of a protein that folds into a given geometrical shape. This method predicts the primary structure of a protein and restores its linear sequence of amino acids in the polypeptide chain using the tertiary structure of a molecule. Unknown amino acids are determined according to the principle of energy minimization. This study represents inverse folding problem as a quadratic optimization problem and uses different relaxation techniques to reduce it to the problem of convex optimizations. Computational experiment compares the quality of these approaches on real protein structures.
△ Less
Submitted 3 January, 2017;
originally announced January 2017.
-
High-contrast Kerr Frequency Combs
Authors:
Ivan S. Grudinin,
Vincent Huet,
Nan Yu,
Michael L. Gorodetsky,
Andrey B. Matsko,
Lute Maleki
Abstract:
Kerr frequency combs with depressed harmonic at the optical pump frequency are theoretically explained and experimentally demonstrated. This result is achieved in a MgF$_2$ photonic belt resonator having reduced density of modes in its spectrum and configured with the add-drop optical couplers. ©2016 California Institute of Technology. Government sponsorship
Kerr frequency combs with depressed harmonic at the optical pump frequency are theoretically explained and experimentally demonstrated. This result is achieved in a MgF$_2$ photonic belt resonator having reduced density of modes in its spectrum and configured with the add-drop optical couplers. ©2016 California Institute of Technology. Government sponsorship
△ Less
Submitted 23 December, 2016; v1 submitted 2 December, 2016;
originally announced December 2016.
-
Fluoride microresonators for mid-IR applications
Authors:
Ivan S. Grudinin,
Kamjou Mansour,
Nan Yu
Abstract:
We study crystalline fluoride microresonators for mid-infrared applications. Whispering gallery mode resonators were fabricated with BaF$_2$, CaF$_2$ and MgF$_2$ crystals. The quality factors were measured at wavelengths of 1.56 μm and 4.58 μm. The impacts of fabrication technique, impurities, multiphonon absorption and surface water are investigated. It is found that MgF2 resonators have room tem…
▽ More
We study crystalline fluoride microresonators for mid-infrared applications. Whispering gallery mode resonators were fabricated with BaF$_2$, CaF$_2$ and MgF$_2$ crystals. The quality factors were measured at wavelengths of 1.56 μm and 4.58 μm. The impacts of fabrication technique, impurities, multiphonon absorption and surface water are investigated. It is found that MgF2 resonators have room temperature Q factor of $8.3\times 10^6$ at wavelength of 4.58 μm, limited by multiphonon absorption.
△ Less
Submitted 1 February, 2016;
originally announced February 2016.
-
Micro--structured crystalline resonators for optical frequency comb generation
Authors:
Ivan S. Grudinin,
Nan Yu
Abstract:
Optical frequency combs have recently been demonstrated in micro--resonators through nonlinear Kerr processes. Investigations in the past few years provided better understanding of micro--combs and showed that spectral span and mode locking are governed by cavity spectrum and dispersion. While various cavities provide unique advantages, dispersion engineering has been reported only for planar wave…
▽ More
Optical frequency combs have recently been demonstrated in micro--resonators through nonlinear Kerr processes. Investigations in the past few years provided better understanding of micro--combs and showed that spectral span and mode locking are governed by cavity spectrum and dispersion. While various cavities provide unique advantages, dispersion engineering has been reported only for planar waveguides. In this Letter, we report a resonator design that combines dispersion control, mode crossing free spectrum, and ultra--high quality factor. We experimentally show that as the dispersion of a MgF2 resonator is flattened, the comb span increases reaching 700 nm with as low as 60 mW pump power at 1560 nm wavelength, corresponding to nearly 2000 lines separated by 46 GHz. The new resonator design may enable efficient low repetition rate coherent octave spanning frequency combs without the need for external broadening, ideal for applications in optical frequency synthesis, metrology, spectroscopy, and communications.
△ Less
Submitted 10 June, 2014;
originally announced June 2014.
-
Frequency combs from crystalline resonators: influence of cavity parameters on comb dynamics
Authors:
Ivan S. Grudinin,
N. Yu
Abstract:
We experimentally study the factors that influence the span in frequency combs derived from the crystalline whispering gallery mode resonators. We observe that cavity dispersion plays an important role in generation of combs by cascaded four wave mixing process. We observed combs from the resonators with anomalous dispersion and nearly zero dispersion at the pump wavelength. In addition, the comb…
▽ More
We experimentally study the factors that influence the span in frequency combs derived from the crystalline whispering gallery mode resonators. We observe that cavity dispersion plays an important role in generation of combs by cascaded four wave mixing process. We observed combs from the resonators with anomalous dispersion and nearly zero dispersion at the pump wavelength. In addition, the comb generation efficiency is found to be affected by the crossing of modes of different families. The influence of Raman gain is discussed as well as the role of cavity diameter and pump power. Copyright 2014 California Institute of Technology. Government sponsorship acknowledged.
△ Less
Submitted 3 June, 2014;
originally announced June 2014.
-
Waveguide couplers for ferroelectric optical resonators
Authors:
Ivan S. Grudinin,
A. Kozhanov,
N. Yu
Abstract:
We report a study of using the same material to fabricate a whispering gallery mode resonator and a coupler. Coupling to high Q whispering gallery modes of the lithium niobate resonator is demonstrated by means of the titanium-doped waveguide. The waveguide coupling approach opens possibilities for simpler and wider practical usage of whispering gallery mode resonators and their integration into o…
▽ More
We report a study of using the same material to fabricate a whispering gallery mode resonator and a coupler. Coupling to high Q whispering gallery modes of the lithium niobate resonator is demonstrated by means of the titanium-doped waveguide. The waveguide coupling approach opens possibilities for simpler and wider practical usage of whispering gallery mode resonators and their integration into optical devices.
△ Less
Submitted 25 April, 2014;
originally announced April 2014.
-
Impact of cavity spectrum on span in microresonator frequency combs
Authors:
Ivan S. Grudinin,
Lukas Baumgartel,
Nan Yu
Abstract:
We experimentally study the factors that limit the span in frequency combs derived from the crystalline whispering gallery mode resonators. We observe that cavity dispersion is the key property that governs the parameters of the combs resulting from cascaded four wave mixing process. Two different regimes of comb generation are observed depending on the precise cavity dispersion behavior at the pu…
▽ More
We experimentally study the factors that limit the span in frequency combs derived from the crystalline whispering gallery mode resonators. We observe that cavity dispersion is the key property that governs the parameters of the combs resulting from cascaded four wave mixing process. Two different regimes of comb generation are observed depending on the precise cavity dispersion behavior at the pump wavelength. In addition, the comb generation efficiency is found to be affected by the crossing of modes of different families. The influence of Raman lasing is discussed.
△ Less
Submitted 17 September, 2013;
originally announced September 2013.
-
Polarization conversion loss in birefringent crystalline resonators
Authors:
Ivan S. Grudinin,
Guoping Lin,
Nan Yu
Abstract:
Whispering gallery modes in birefringent crystalline resonators are investigated. We experimentally investigate the XY--cut resonators made with LiNbO$_3$, LiTaO$_3$ and BBO and observe strong influence of the resonator's shape and birefringence on the quality factor of the extraordinary polarized modes. We show that extraordinary modes can have lower Q and even be suppressed due to polarization c…
▽ More
Whispering gallery modes in birefringent crystalline resonators are investigated. We experimentally investigate the XY--cut resonators made with LiNbO$_3$, LiTaO$_3$ and BBO and observe strong influence of the resonator's shape and birefringence on the quality factor of the extraordinary polarized modes. We show that extraordinary modes can have lower Q and even be suppressed due to polarization conversion loss. The ordinary ray modes retain the high Q due to inhibited reflection phenomenon.
△ Less
Submitted 3 April, 2013;
originally announced April 2013.
-
Frequency comb from a microresonator with engineered spectrum
Authors:
Ivan S. Grudinin,
Lukas Baumgartel,
Nan Yu
Abstract:
We demonstrate that by varying the ratio between the linewidth and dispersion of a whispering gallery mode resonator we are able to control the number N of free spectral ranges separating the first generated comb sidebands from the pump. We observed combs with N=19 and N=1. For the comb with N=1 we have achieved a span of over 200 nm using a 0.4 mm MgF2 resonator with a pump of 50 mW, which is a f…
▽ More
We demonstrate that by varying the ratio between the linewidth and dispersion of a whispering gallery mode resonator we are able to control the number N of free spectral ranges separating the first generated comb sidebands from the pump. We observed combs with N=19 and N=1. For the comb with N=1 we have achieved a span of over 200 nm using a 0.4 mm MgF2 resonator with a pump of 50 mW, which is a factor of 10 lower than previously reported.
△ Less
Submitted 6 February, 2012;
originally announced February 2012.
-
Temperature measurement and stabilization in a birefringent whispering gallery resonator
Authors:
D. V. Strekalov,
R. J. Thompson,
L. M. Baumgartel,
I. S. Grudinin,
N. Yu
Abstract:
Temperature measurement with nano-Kelvin resolution is demonstrated at room temperature, based on the thermal dependence of an optical crystal anisotropy in a high quality whispering gallery resonator. As the resonator's TE and TM modes frequencies have different temperature coefficients, their differential shift provides a sensitive measurement of the temperature variation, which is used for acti…
▽ More
Temperature measurement with nano-Kelvin resolution is demonstrated at room temperature, based on the thermal dependence of an optical crystal anisotropy in a high quality whispering gallery resonator. As the resonator's TE and TM modes frequencies have different temperature coefficients, their differential shift provides a sensitive measurement of the temperature variation, which is used for active stabilization of the temperature.
△ Less
Submitted 31 May, 2011;
originally announced May 2011.
-
Phonon laser action in a tunable, two-level photonic molecule
Authors:
Ivan S. Grudinin,
O. Painter,
Kerry J. Vahala
Abstract:
The phonon analog of an optical laser has long been a subject of interest. We demonstrate a compound microcavity system, coupled to a radio-frequency mechanical mode, that operates in close analogy to a two-level laser system. An inversion produces gain, causing phonon laser action above a pump power threshold of around 50 $μ$W. The device features a continuously tunable, gain spectrum to select…
▽ More
The phonon analog of an optical laser has long been a subject of interest. We demonstrate a compound microcavity system, coupled to a radio-frequency mechanical mode, that operates in close analogy to a two-level laser system. An inversion produces gain, causing phonon laser action above a pump power threshold of around 50 $μ$W. The device features a continuously tunable, gain spectrum to selectively amplify mechanical modes from radio frequency to microwave rates. Viewed as a Brillouin process, the system accesses a regime in which the phonon plays what has traditionally been the role of the Stokes wave. For this reason, it should also be possible to controllably switch between phonon and photon laser regimes. Cooling of the mechanical mode is also possible.
△ Less
Submitted 29 July, 2009;
originally announced July 2009.
-
Brillouin Lasing with a CaF_2 Whispering Gallery Mode Resonator
Authors:
Ivan S. Grudinin,
Andrey B. Matsko,
Lute Maleki
Abstract:
Stimulated Brillouin scattering with both pump and Stokes beams in resonance with whispering gallery modes of an ultra high Q CaF_2 resonator is demonstrated for the first time. The resonator is pumped with 1064 nm light and has a Brillouin lasing threshold of 3.5 microwatt. Potential applications include optical generation of microwaves and sensitive gyros.
Stimulated Brillouin scattering with both pump and Stokes beams in resonance with whispering gallery modes of an ultra high Q CaF_2 resonator is demonstrated for the first time. The resonator is pumped with 1064 nm light and has a Brillouin lasing threshold of 3.5 microwatt. Potential applications include optical generation of microwaves and sensitive gyros.
△ Less
Submitted 7 May, 2008;
originally announced May 2008.
-
Enhanced photothermal displacement spectroscopy for thin-film characterization using a Fabry-Perot resonator
Authors:
Eric D. Black,
Ivan S. Grudinin,
Shanti R. Rao,
Kenneth G. Libbrecht
Abstract:
We have developed a new technique for photothermal displacement spectroscopy that is potentially orders of magnitude more sensitive than conventional methods. We use a single Fabry-Perot resonator to enhance both the intensity of the pump beam and the sensitivity of the probe beam. The result is an enhancement of the response of the instrument by a factor proportional to the square of the finess…
▽ More
We have developed a new technique for photothermal displacement spectroscopy that is potentially orders of magnitude more sensitive than conventional methods. We use a single Fabry-Perot resonator to enhance both the intensity of the pump beam and the sensitivity of the probe beam. The result is an enhancement of the response of the instrument by a factor proportional to the square of the finesse of the cavity over conventional interferometric measurements.
In this paper we present a description of the technique, and we discuss how the properties of thin films can be deduced from the photothermal response. As an example of the technique, we report a measurement of the thermal properties of a multilayer dielectric mirror similar to those used in interferometric gravitational wave detectors.
△ Less
Submitted 5 January, 2004; v1 submitted 8 October, 2003;
originally announced October 2003.
-
Fundamental thermal fluctuations in microspheres
Authors:
M. L. Gorodetsky,
I. S. Grudinin
Abstract:
We present theoretical analysis and the results of measurements of fundamental thermorefractive fluctuations in microspheres. Experimentally measured noise spectra are consistent with the theoretical model.
We present theoretical analysis and the results of measurements of fundamental thermorefractive fluctuations in microspheres. Experimentally measured noise spectra are consistent with the theoretical model.
△ Less
Submitted 2 April, 2003;
originally announced April 2003.