Igor Fedorov

Igor Fedorov

Auto-generated profile, based on publicly available information

Igor Fedorov

Meta, Menlo Park, CA, USA

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 April 2018 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
ANAlican Nalci
Alican Nalci
IFIgor Fedorov
Igor Fedorov
MAMaher Al-Shoukairi
Maher Al-Shoukairi
TLThomas T. Liu
Thomas T. Liu
BRBhaskar D. Rao
Bhaskar D. Rao
In this paper, we develop a Bayesian evidence maximization framework to solve the sparse non-negative least squares (S-NNLS) problem. We introduce a family of probability densities referred to as the rectified Gaussian scale mixture (R-GSM) to model the sparsity enforcing prior distribution for the solution. The R-GSM prior encompasses a variety of heavy-tailed densities such as the rectified Laplacian and rectified Student's t-distributions with a proper choice of the mixing density. We utilize the hierarchical representation induced by the R-GSM prior and develop an evidence maximization framework based on the expectation-maximization (EM) algorithm. Using the EM based method, we estimate the hyper-parameters and obtain a point estimate for the solution. We refer to the proposed method as rectified sparse Bayesian learning (R-SBL). We provide four R-SBL variants that offer a range of options for computational complexity and the quality of the E-step computation. These methods include the Markov chain Monte Carlo EM, linear minimum mean-square-error estimation, approximate message passing, and a diagonal approximation. Using numerical experiments, we show that the proposed R-SBL method outperforms existing S-NNLS solvers in terms of both signal and support recovery performance, and is also very robust against the structure of the design matrix.

6 January 2018 in Signal Processing
Verified
Journal Article
Closed access
IFIgor Fedorov
Igor Fedorov
ANAlican Nalci
Alican Nalci
RGRitwik Giri
Ritwik Giri
BRBhaskar D. Rao
Bhaskar D. Rao
TNTruong Q. Nguyen
Truong Q. Nguyen
HGHarinath Garudadri
Harinath Garudadri

1 March 2020
Verified
Conference Paper
Closed access
YKYuri N. Korkishko
Yuri N. Korkishko
VFVyacheslav A. Fedorov
Vyacheslav A. Fedorov
VPViktor E. Prilutskiy
Viktor E. Prilutskiy
VPVladimir G. Ponomarev
Vladimir G. Ponomarev
SPStanislav V. Prilutskiy
Stanislav V. Prilutskiy
IMIvan V. Morev
Ivan V. Morev
At present time interferometric fiber-optic gyroscopes (IFOG, FOG) are widely used in inertial navigation systems (INS), and in wide range of applications have replaced its well-established main competitor ring laser gyroscopes (RLG). Recently, in order to cover the mass-market applications spectrum requiring low-cost and compact inertial sensor yet as much precise as it can be, RPC Optolink has launched new IFOG-based product: ultra-compact navigation-grade inertial measurement unit IMU400, its SWaP properties are: 80×95×62 mm, 0.7 kg, 0.5 l, ≤7 W. The aim of the current work was the production of pilot IMU400 devices batch and the estimation of the performance of IMU with direct approach and also with strapdown inertial navigation systems (SINS) simulation methods, which by sense is indirect way of performance observation. Main IMU400 Gyro and Accelerometer parameters are: Angle Random Walk (ARW) = 0.007 °/√hour, Bias Instability (BI) = 0.01°/h; Velocity Random Walk (VRW) = 40µg/√Hz, BI = 6µg. SINS expected performance (max): heading 0.2°×sec(lat) (1σ, 10 min align time).

13 May 2024
Verified
Conference Paper
Open access
WWWei Wen
Wei Wen
KLKuang-Hung Liu
Kuang-Hung Liu
IFIgor Fedorov
Igor Fedorov
XZXin Zhang
Xin Zhang
HYHang Yin
Hang Yin
WCWen-Yen Chen
Wen-Yen Chen
Neural Architecture Search (NAS) has demonstrated its efficacy in computer vision and potential for ranking systems. However, prior work focused on academic problems, which are evaluated at small scale under well-controlled fixed baselines. In industry system, such as ranking system in Meta, it is unclear whether NAS algorithms from the literature can outperform production baselines because of: (1) scale - Meta's ranking systems serve billions of users, (2) strong baselines - the baselines are production models optimized by hundreds to thousands of world-class engineers for years since the rise of deep learning, (3) dynamic baselines - engineers may have established new and stronger baselines during NAS search, and (4) efficiency - the search pipeline must yield results quickly in alignment with the productionization life cycle. In this paper, we present Rankitect, a NAS software framework for ranking systems at Meta. Rankitect seeks to build brand new architectures by composing low level building blocks from scratch. Rankitect implements and improves state-of-the-art (SOTA) NAS methods for comprehensive and fair comparison under the same search space, including sampling-based NAS, one-shot NAS, and Differentiable NAS (DNAS). We evaluate Rankitect by comparing to multiple production ranking models at Meta. We find that Rankitect can discover new models from scratch achieving competitive trade-off between Normalized Entropy loss and FLOPs. When utilizing search space designed by engineers, Rankitect can generate better models than engineers, achieving positive offline evaluation and online A/B test at Meta scale.

24 August 2024
Verified
Conference Paper
Open access
HZHua Zheng
Hua Zheng
KLKuang-Hung Liu
Kuang-Hung Liu
IFIgor Fedorov
Igor Fedorov
XZXin Zhang
Xin Zhang
WCWen-Yen Chen
Wen-Yen Chen
WWWei Wen
Wei Wen
Neural Architecture Search (NAS) has become a widely used tool for automating neural network design. While one-shot NAS methods have successfully reduced computational requirements, they often require extensive training. On the other hand, zero-shot NAS utilizes training-free proxies to evaluate a candidate architecture's test performance but has two limitations: (1) inability to use the information gained as a network improves with training and (2) unreliable performance, particularly in complex domains like RecSys, due to the multi-modal data inputs and complex architecture configurations. To synthesize the benefits of both methods, we introduce a "sub-one-shot" paradigm that serves as a bridge between zero-shot and one-shot NAS. In sub-one-shot NAS, the supernet is trained using only a small subset of the training data, a phase we refer to as "warm-up." Within this framework, we present SiGeo, a proxy founded on a novel theoretical framework that connects the supernet warm-up with the efficacy of the proxy. Extensive experiments have consistently shown that SiGeo, when properly warmed up, surpasses state-of-the-art NAS proxies in many established NAS benchmarks in the computer vision domain. Furthermore, when tested on recommendation system benchmarks, SiGeo demonstrates its ability to match the performance of state-of-the-art weight-sharing one-shot NAS methods while significantly reducing computational costs by approximately 60%.

22 July 2017 in arXiv
Verified
Preprint
Open access
PDF file
IFIgor Fedorov
Igor Fedorov
ANAlican Nalci
Alican Nalci
RGRitwik Giri
Ritwik Giri
BRBhaskar D. Rao
Bhaskar D. Rao
TNTruong Q. Nguyen
Truong Q. Nguyen
HGHarinath Garudadri
Harinath Garudadri
We study the sparse non-negative least squares (S-NNLS) problem. S-NNLS occurs naturally in a wide variety of applications where an unknown, non-negative quantity must be recovered from linear measurements. We present a unified framework for S-NNLS based on a rectified power exponential scale mixture prior on the sparse codes. We show that the proposed framework encompasses a large class of S-NNLS algorithms and provide a computationally efficient inference procedure based on multiplicative update rules. Such update rules are convenient for solving large sets of S-NNLS problems simultaneously, which is required in contexts like sparse non-negative matrix factorization (S-NMF). We provide theoretical justification for the proposed approach by showing that the local minima of the objective function being optimized are sparse and the S-NNLS algorithms presented are guaranteed to converge to a set of stationary points of the objective function. We then extend our framework to S-NMF, showing that our framework leads to many well known S-NMF algorithms under specific choices of prior and providing a guarantee that a popular subclass of the proposed algorithms converge to a set of stationary points of the objective function. Finally, we study the performance of the proposed approaches on synthetic and real-world data.