Mountain View, California, United States
Contact Info
6K followers
500+ connections
About
Activity
-
A few weeks back, I handed over my Google badge. Reflecting on my time at Google, I feel extremely lucky for the many opportunities I have been…
A few weeks back, I handed over my Google badge. Reflecting on my time at Google, I feel extremely lucky for the many opportunities I have been…
Liked by Fei Xia
-
New from FAIR: An Introduction to Vision-Language Modeling. Paper ➡️ https://go.fb.me/ncjj6t This guide covers how VLMs work, how to train them and…
New from FAIR: An Introduction to Vision-Language Modeling. Paper ➡️ https://go.fb.me/ncjj6t This guide covers how VLMs work, how to train them and…
Liked by Fei Xia
-
After an incredible eight-year journey working in Nuro and then Waymo, as I decided to take a break on autonomous driving, the past week marked my…
After an incredible eight-year journey working in Nuro and then Waymo, as I decided to take a break on autonomous driving, the past week marked my…
Liked by Fei Xia
Experience & Education
Licenses & Certifications
Publications
-
Block-coordinate Frank-Wolfe Optimization for Counting Objects in Images
Neural Information Processing Systems (NeurIPS) Optimization Workshop
We develop an optimization method to count objects in images. To avoid the detection of individual objects, which is computationally expensive and relies heavily on image quality, we model the object density as a linear transformation of each pixel feature, and obtain the object count by integrating density over the image. Learning such linear transformation is formulated as the minimization of a regularized quadratic function. Solving this optimization problem is highly nontrivial because it…
We develop an optimization method to count objects in images. To avoid the detection of individual objects, which is computationally expensive and relies heavily on image quality, we model the object density as a linear transformation of each pixel feature, and obtain the object count by integrating density over the image. Learning such linear transformation is formulated as the minimization of a regularized quadratic function. Solving this optimization problem is highly nontrivial because it has exponentially large number of constraints. To cope with this challenge, inspired by the structural support vector machine (SVM), we explore the Block-Coordinate Frank-Wolfe (BCFW) algorithm, which is a state-of-the-art algorithm to solve structural SVM. However, BCFW cannot be directly applied to our problem. We derive the dual of our optimization problem and solve it by BCFW with modifications. Experiments show that BCFW solves our problem with lower iteration cost, faster convergence, and decent error rate.
Other authorsSee publication -
Max-margin Latent Feature Relational Models for Entity-Attribute Networks
International Joint Conference on Neural Networks (IJCNN)
Link prediction is a fundamental task in statistical analysis of network data. Though much research has concentrated on predicting entity-entity relationships in homogeneous networks, it has attracted increasing attentions to predict relationships in heterogeneous networks, which consist of multiple types of nodes and relational links. Existing work on heterogeneous network link prediction mainly focuses on using input features that are explicitly extracted by humans. This paper presents…
Link prediction is a fundamental task in statistical analysis of network data. Though much research has concentrated on predicting entity-entity relationships in homogeneous networks, it has attracted increasing attentions to predict relationships in heterogeneous networks, which consist of multiple types of nodes and relational links. Existing work on heterogeneous network link prediction mainly focuses on using input features that are explicitly extracted by humans. This paper presents an
approach to automatically learn latent features from partially observed heterogeneous networks, with a particular focus on entity-attribute networks (EANs), and making predictions for
unseen pairs. To make the latent features discriminative, we adopt the max-margin idea under the framework of maximum entropy discrimination (MED). Our maximum entropy discrimination joint relational model (MED-JRM) can jointly predict entity-entity relationships as well as the missing attributes of entities in EANs. Experimental results on several real networks demonstrate
that our model has improved performance over state-of-the-art homogeneous and heterogeneous network link prediction algorithms.Other authorsSee publication -
Parameter Server for Distributed Machine Learning
Neural Information Processing Systems (NeurIPS Workshop)
We propose a parameter server framework to solve distributed machine learning problems. Both data and workload are distributed into client nodes, while server nodes maintain globally shared parameters, which are represented as sparse vectors and matrices. The framework manages asynchronous data communications between clients and servers. Flexible consistency models, elastic scalability and fault tolerance are supported by this framework. We present algorithms and theoretical analysis for…
We propose a parameter server framework to solve distributed machine learning problems. Both data and workload are distributed into client nodes, while server nodes maintain globally shared parameters, which are represented as sparse vectors and matrices. The framework manages asynchronous data communications between clients and servers. Flexible consistency models, elastic scalability and fault tolerance are supported by this framework. We present algorithms and theoretical analysis for challenging nonconvex and nonsmooth problems. To demonstrate the scalability of the proposed framework, we show experimental results on real data with billions of parameters.
Other authorsSee publication -
Generalized Relational Topic Models with Data Augmentation
International Joint Conferences on Artificial Intelligence (IJCAI)
Relational topic models have shown promise on analyzing document network structures and discovering latent topic representations. This paper presents three extensions: 1) unlike the common link likelihood with a diagonal weight matrix that allows the-same-topic interactions only, we generalize it to use a full weight matrix that captures all pairwise topic interactions and is applicable to asymmetric networks; 2) instead of doing standard Bayesian inference, we perform regularized Bayesian…
Relational topic models have shown promise on analyzing document network structures and discovering latent topic representations. This paper presents three extensions: 1) unlike the common link likelihood with a diagonal weight matrix that allows the-same-topic interactions only, we generalize it to use a full weight matrix that captures all pairwise topic interactions and is applicable to asymmetric networks; 2) instead of doing standard Bayesian inference, we perform regularized Bayesian inference with a regularization parameter to deal with the imbalanced link structure issue in common real networks; and 3) instead of doing variational approximation with strict mean-field assumptions, we present a collapsed Gibbs sampling algorithm for the generalized relational topic models without making restricting assumptions. Experimental results demonstrate the significance of these extensions on improving the prediction performance, and the time efficiency can be dramatically improved with a simple fast approximation method.
Other authorsSee publication -
Discriminative Relational Topic Models
Submitted to Pattern Analysis and Machine Intelligence (PAMI)
Many scientific and engineering fields involve analyzing network data. For document networks, relational topic models (RTMs) provide a probabilistic generative process to describe both the link structure and document contents, and they have shown promise on predicting network structures and discovering latent topic representations. However, existing RTMs have limitations in both the restricted model expressiveness and incapability of dealing with imbalanced network data. To expand the scope and…
Many scientific and engineering fields involve analyzing network data. For document networks, relational topic models (RTMs) provide a probabilistic generative process to describe both the link structure and document contents, and they have shown promise on predicting network structures and discovering latent topic representations. However, existing RTMs have limitations in both the restricted model expressiveness and incapability of dealing with imbalanced network data. To expand the scope and improve the inference accuracy of RTMs, this paper presents three extensions: 1) unlike the common link likelihood with a diagonal weight matrix that allows the-same-topic interactions only, we generalize it to use a full weight matrix that captures all pairwise topic interactions and is applicable to asymmetric networks; 2) instead of doing standard Bayesian inference, we perform regularized Bayesian inference (RegBayes) with a regularization parameter to deal with the imbalanced link structure issue in common real networks and improve the discriminative ability of learned latent representations; and 3) instead of doing variational approximation with strict mean-field assumptions, we present collapsed Gibbs sampling algorithms for the generalized relational topic models by exploring data augmentation without making restricting assumptions. Under the generic RegBayes framework, we carefully investigate two popular discriminative loss functions, namely, the logistic log-loss and the max-margin hinge loss. Experimental results on several real network datasets demonstrate the significance of these extensions on improving the prediction performance, and the time efficiency can be dramatically improved with a simple fast approximation method.
Other authorsSee publication
Projects
-
Location Recommendation Tool
- Present
1. Aim to deliver a geographic information analysis and recommendation tool for PNC Bank
2. Mine information from Foursquare and Twitter data
3. Develop a website under Django framework and with spatial database (PostgreSQL + PostGIS)Other creators -
-
Search Engine System
-
1. Implemented a search engine system with multiple retrieval algorithms, such as Boolean Retrieval model, BM25 and Indri, to retrieve Wikipedia documents indexed by LUCENE
2. Optimized the system’s performance through using various techniques, such as multiple representation model, sequential dependency model and pseudo relevance feedback
Languages
-
Chinese
Native or bilingual proficiency
-
English
Full professional proficiency
More activity by Fei
-
Today concludes my journey with Google, a chapter that began immediately after my graduation six years and eight months ago. Working alongside a team…
Today concludes my journey with Google, a chapter that began immediately after my graduation six years and eight months ago. Working alongside a team…
Liked by Fei Xia
-
#waymo will be at #iccv2023! I will be there also, if you want to meet up let me know. Come check out our workshop talks (see below)! In addition…
#waymo will be at #iccv2023! I will be there also, if you want to meet up let me know. Come check out our workshop talks (see below)! In addition…
Liked by Fei Xia
-
Seeking to hire passionate researchers/engineers to work on developing the most advanced multimodal (vision-language) models. Candidates with…
Seeking to hire passionate researchers/engineers to work on developing the most advanced multimodal (vision-language) models. Candidates with…
Liked by Fei Xia
-
It has been an incredible five year Journey for me at XPENG. Together, we built the best-in-class autonomous driving technologies in mass production…
It has been an incredible five year Journey for me at XPENG. Together, we built the best-in-class autonomous driving technologies in mass production…
Liked by Fei Xia
-
Late last year at #corl2022, I gave a talk highlighting some of our recent #waymo #research on behavior modeling topics. (For Perception-focused…
Late last year at #corl2022, I gave a talk highlighting some of our recent #waymo #research on behavior modeling topics. (For Perception-focused…
Liked by Fei Xia
-
#waymo has a strong presence in this year's #eccv2022, where we will be presenting 8 papers. If you want to learn more about our recent 3D…
#waymo has a strong presence in this year's #eccv2022, where we will be presenting 8 papers. If you want to learn more about our recent 3D…
Liked by Fei Xia
-
4 HOURS OF LOFI BEATS AND WAYMO AUTONOMOUS DRIVING TO STUDY AND RELAX TO #WEAREWAYMO
4 HOURS OF LOFI BEATS AND WAYMO AUTONOMOUS DRIVING TO STUDY AND RELAX TO #WEAREWAYMO
Liked by Fei Xia
-
Hey! Jyh-Jing Hwang, myself, and several other #waymo coworkers of mine are going to be at #eccv2022 this coming week in Israel to present several…
Hey! Jyh-Jing Hwang, myself, and several other #waymo coworkers of mine are going to be at #eccv2022 this coming week in Israel to present several…
Liked by Fei Xia
-
I am super excited to start a new position at Chef Robotics! It’s always been a dream for me to work at an early-stage robotics startup. Chef…
I am super excited to start a new position at Chef Robotics! It’s always been a dream for me to work at an early-stage robotics startup. Chef…
Liked by Fei Xia
-
Interactive behavior prediction might be one of the most challenging and urgent problems in autonomous driving. Glad that we have proposed a…
Interactive behavior prediction might be one of the most challenging and urgent problems in autonomous driving. Glad that we have proposed a…
Liked by Fei Xia
Other similar profiles
-
Chen Wu
Connect -
RUOYAN LIU
Connect -
Kun Huang
Connect -
Bill Z.
Connect -
Yunhan Ning
Senior Software Engineer at Meituan
Connect -
Chang Gao
Connect -
Haotian Wu
Connect -
Yimeng Zhang
Connect -
Yuning Chai
Connect -
Jian Han
Vice President Of Engineering/GM of Online Delievery Platform at Meituan-Dianping
Connect
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Fei Xia in United States
-
Fei Xia
Senior Research Scientist at Google DeepMind, Working on Robotics
-
Fei Xia
-
Fei Xia
-
Fei Xia
Fei Xia, soprano
15 others named Fei Xia in United States are on LinkedIn
See others named Fei Xia