Welcome! You're the
th visitor since 05/25/2021.
Last Update: May, 23th, 2022.
Zhengqi(Drago) Dong
Boston Universitydong760 {at} bu {dot} edu
drago1234
Intro: Hi guys, I am Zhengqi, and you can call me Drago as well (It's a nickname that I picked up from a moive when I was first time travel to US. You can read more here, if you are interested). Currently, I am a software engineer intern at Yrobot. For my academic background, I earned my Computer Science and Engineering bachelor's degree @ The Ohio State University (with a specialization in AI and a minor in Statistics), and a Master degree in Robotic and Autonomous System @ Boston University. I am highly interested in the area of AI/ML/DL, HPC, and with becoming a professional robotics engineer as my long-term goal in mind! Love to connect and grow with buddies sharing the same interests!
Objective: looking for an Full Time New Grad SDE position based on United States, expected starting date: Jan - Mar, 2023.
News
2022-July: If I couldn't find a good opportunity here by the end of Dec 2022, likely I will be joining Huawei.2022-May: I will be an intern for Y-Robot company (A robotic startup for wearable devices, read more here: src1, src2, src3, src4) this summer as a Software Engineer, at Cambridge, MA.
2022-Jan: I will be a Course Assistant for DS/CS-519 -- Spark! Software Engineering X-Lab Practicum, helping Prof. Langdon White for general course errand, and holding office hour 10-12pm Thursday and Friday each week!
2021-Sep: I will start my graduate study at Boston University toward Master of Science in Robotics and Autonomous System.
2021-April: I will be giving a presentation at URF(Undergraduate Research Forum) on 4/1 (More details about URF at Link) and another one at Denman Undergraduate Research Forum on Monday, April 5th, 2021, 4:00 pm–5:30 pm – Engineering: Physical Sciences Panel, register at here Link (More details about Denman Undergraduate Research Forum at Link)
2020-Aug: Employed as Instruction Assistant for CSE 3461: Computer Networking and Internet Technologies
2020-Aug: I completed a Working Behavior test that is supported by Plum. It was a behavior test that requested by Bloomberg Recruit Team. And, I think the result can say something about me!
Education
Boston University (Fun fact: Ranked #1 in the top 100 colleges with the Hardest Final Exam)
Boston,
Massachusetts, U.S.A
Master of
Science in Robotics & Autonomous Systems
GPA
3.93/4.0
(Degree Advice Report)
Aug. 2021 - Dec. 2022 (Expected)
The Ohio State University
Columbus,
OH, U.S.A
Bachelor of
Computer Science and Engineering (Minor: Statistics & Specialization: AI)
Honor: Graduated with Honor in Engineering (29/317), and Honor Research Distinction (3/317).
Thesis: Image-Based Plant Leaf Disease Recognition with InceptionV3 Network (Source)
GPA 3.65/4.0 (Audit Course Report, Course History)
Thesis: Image-Based Plant Leaf Disease Recognition with InceptionV3 Network (Source)
GPA 3.65/4.0 (Audit Course Report, Course History)
Aug. 2017 - May. 2021
University of Dayton
Dayton,
OH, U.S.A
Bachelor of
Mecahnical Engineering (Transfer out)
GPA
3.87/4.0
Sept. 2015 - July 2017
Foreign Language School Attached to Guangxi Normal University (广西师范大学附属外国语学校)
Guilin,
Guangxi, P.R.China
Sept. 2013 - June 2015
Projects
2022 Fall--BU ME570, Motion Planning Final Project: A Deep Learning Based Approach for Foot Placement Prediction with Attention Mechanism [Course Syllabus]
Abstract: Predicting the next foot placement of human walking or running activity is an important task for developing a comfortable and intelligent walking aid robot. The previous study mainly focuses on designing the walking aid robot with better mechanical support or adding some traditional machining learning approaches that require extensive manual feature selection to generate the foot placement probability grid map. Those approaches can be difficult to generalize to various complicated scenarios in human daily life activities. Therefore, an attention mechanism with several state-of-the-art neural network techniques was investigated for improving the generalization ability and accuracy of the foot placement prediction task. In this study, we implemented and compared four different neural network architectures for foot placement prediction, and the experimental result are obtained and showed a clear improvement in the effectiveness of adding an attention mechanism to the novel network architecture that only consists of Long Short-Term Memory (LSTM) layer or Bidirectional LSTM (Bi-LSTM) layer. An experimental comparison found that the MAE, MSE, and RMSE prediction error has reduced by about 8.64%, 16.97%, and 8.88% after adding the attention layer meanwhile this performance does not correlate with the model size. Besides, our result also shows that simply replacing the LSTM with the Bi-LSTM layer would not improve the performance and can be harmful instead.
2022Spring--BU EC523, Deep Learning Project: Automatic Waste Detection on ZeroWaste [Course Syllabus]
Summary: Increased air, water, soil and sea pollution is still an big problem today. According to
the data forecasted by World Bank Group at
2018(p24), 2.59 billion tonnes of waste will be generated annually by 2030, and this number will be expected to reach 3.40 billion tonnes by 2050. In addition, based on the data reported by States
Environmental Protection Agency, there is less than 35% of recyclable waste is actually being recycled in the US. Due to high cost of waste management($50-$100 per tonnes), inadequate waste collection and uncontrolled dumping or burning of solid waste are still an unfortunated reality, polluting the air, waster, and soil in the world. Therefore, in this project, we are aiming at pushing on the barriers of automatic waste detection and sorting and try to experiment different state-of-the-art detection model (both single-stage and two-stage detector) on the Zero Waste dataset, the largest public waste detection dataset. At the end, three detection models were tested with some customized pre-processing and data augmentation, and their performance all exceeds the baseline shown in the original paper that was published three monthes ago. Among them, YOLOR has achieved the best performance over all evaluation criteria and has outperformed TridentNet greatly, 2.57x higher on AP and 5.92x higher on AP_small. The detailed benchmark and explanations are shown in the report!
2022Spring--BU CS585, Image Processing Challenge: Monocular camera based Visual Odometry(or Localization) [Course Syllabus]
2022Spring--BU EC527 Final Project: CNN Optimization(Single-core, Multi-core CPU, and GPU) [Course Syllabus]
2021 Fall--BU ME571 Medical Robotic Final Project: 3D Printed Prosthetic Robotic Grasper [Course Syllabus]
2021Spring--CSE Honor Research Project: Image-Based Plant Leaf Disease Detection using
InceptionV3
2021Spring--CSE 5901.01: Scaling Distributed DNN Training for Large Images(continued with last semester
project)
2020 Fall--CSE 5249 Research Project: Scaling Distributed DNN Training for Large Images, [OSU HiDL Lab]
.png)
2020Fall--CSE 5914 AI Capstone: Filmpedia -- A movie recommendation website
2020Fall--CSE 5525 Speech and Language Processing Final Project, Hybrid Filtering Recommender System [Course Website]
Summary: In this project, we proposed a hybrid filtering movie recommendation system that combines the idea of content-based filtering and collaborative filtering and integrated the representation learned from multiple sources. While conducting the experiments, we explored the performance of text embedding methods by using it as our text-embedding module, and designed unique metapaths and a novel rating-aware sampling policy for graph embedding. Leveraging both text and graph contextual information, the system further improved the quality of recommendation system compared to methods using single information source.
Summary: In this project, we proposed a hybrid filtering movie recommendation system that combines the idea of content-based filtering and collaborative filtering and integrated the representation learned from multiple sources. While conducting the experiments, we explored the performance of text embedding methods by using it as our text-embedding module, and designed unique metapaths and a novel rating-aware sampling policy for graph embedding. Leveraging both text and graph contextual information, the system further improved the quality of recommendation system compared to methods using single information source.
Professional Experiences
-
Software Engineer Intern, Yrobot lnc, Boston, MA, USA, June 2022 - Sep 2022
- Designed and developed a File Transferring Simulator for company’s embedded system of wearable devices.
- Wrote C++ code for dev board and Python code for client endpoint to establish a communication channel via TCP and X/Y/ZMODEM protocol, used CMake to build system and GTest for unit testing.
- Implemented a Bayesian inference-based probability distribution model for foot placement prediction with Python
-
Software Developer Intern, BU Spark!, Boston, MA, USA, Sep 2021 - Jan 2022
- Created a website that loads mutual aid resources from Postgres database, then displays all food resources and mutual aid locations around Greater Boston area in an interactive map by using mapbox API.
- Designed and developed the front-end in Gatsby to improve user experience by adding multi-language feature.
- Deployed frontend via GitHub Pages with https secure access, and utilized Docker Compose to containerize back-end application, then deployed on AWS EC2 instance, and secured the communication between frond-end and backed with TLS/SSL certificate.
-
Student Instructional Assistant, Ohio State University, Columbus, OH, Aug 2020 - Dec 2020
- CSE 3461: Computer Networking and Internet Technologies (Fall 2020)
- Grade students' homework and report basic statistics and common mistakes to the professor.
- Answer students' questions by holding office hours and replying to emails.
-
WebMaster, Student Association of Graduate Engineers (SAGE) at Boston University, Aug 2021 -
Present
- Coodinated with other e-board members to plan and organize two annual events (whale watching and nutcracker)
- Routinely updated and maintained SAGE's website, including event news, student activities, post, and relevant school information, by using WordPress Content Management System (CMS).
Award and Course Certification
- Google -- Tensorflow Certification (Awarded in Dec 2020. You can find me on TensorFlow Certificate Network)
- Coursera -- TensorFlow in Practice Specialization (Awarded in Aug 2020)
- IEEE -- Sensoors for Autonomous Vehicles (Awarded in Seq 2019)
Favorite Quote (Wisdom of the Past)
- Don't waste time dreaming of great far-away opportuniies; start to work and improve yourself right now. Someday a small opportuniies will come, and those are often the beginning of great and amazing things.
- "运筹帷幄之中，决胜千里之外.” ― 西汉·司马迁
- "A small group of dedicated people can change the world!" --A friend of mine
- "Imagination is more important than knowledge." --Alber Einstein
- "If I had an hour to solve a problem I'd spend 55 minutes thinking about the problem and five minutes thinking about solutions." --Alber Einstein
- "You MULTIPLY your time by giving yourself the EMOTIONAL PERMISSION to spend time on things TODAY that will give you more time TOMORROW" --Rory Vaden
- Get the most important thing done at first! -- "Organize Tommorrow Today"
- "I am not afraid of a person who knows 10000 kicks. But I am afraid of a person who knows one kick but practices it for 10000 times." --Bruce Lee
- If you don't measure, you don't know; If you don't profile, you can't tell where it spent.
- "Premature Optimization is the root of all evil." --Knuth
- "When he talks, people listen"-- Warren Buffet
- "Be fearful when others are greedy and greedy when others are fearful." -- Warren Buffet
- "20% of your activitie will account for 80% of your results."" -- Pareto principle
- "What doesn't kill you makes you stronger." -- Friedrich Nietzsche
- “世界上只有一种真正的英雄主义，那就是在认清生活的真相后依然热爱生活。” --罗曼·罗兰 《米开朗琪罗》
- "Low Expectation + Luckiness = Happiness"
- “弱水三千，我只取一瓢" -- 佛说
- "Don't see what they want you to see, you got to open your mind to the possibilities!" -- Bumi
- "All good things are worth waiting for and worth fighting for.” ― Susan Gale
- "我命由我不由天！.” ― Deep Memory
- "Never sacrifice the long-term value for short-term results." -- Amazon Leadership Principles
- "Do not waste time dreaming of great far-away opportuniies; start to work and improve yourself right now. Someday, a small opportunity will come, and those are of the beginning of great and amazing things." -- Unknown
- "Get the most important thing done." -- Unknown
- "Do the most important thing first each day" -- Unknown
Miscellaneous
-
Note: This website is still under development. I am consistently working on the Front-end design of different components, background themes, and try to integrate more meaningul functionality and features to make it looks more professional and easy to navigate! (Search bar is not implemented, please bear with me!)
-
Note: I am using the Jekyll to build this website, but there is another one called Hugo if you interest to know more, click this link here, and here is the Lanyon theme for Hugo.
-
If you have any good suggestion or comments, you can welcome to enter them in the discussion panel below, and I will reponse to you whenever I can!
- TO-DO: 1) Changes Tags to Tag cloud-- an example, 2) Add an login portal for backend data management. 3) Add a portal that can connect to my photo sharing and movie recommendation applications (Add all those graduation photots to your photo sharing app, if time allows!). 4) Design the home page for my knowledge bank, and start thinking about what component I want to put there, and what is the best way to arrange/organize them. 5) Finish the editing of MPI tutorial, and other 3 labs related to Operating System. 6) Consistently update those summary sheet that I created for those coding algorithms.
