Projects

A General Method For Automatic Discovery of Powerful Interactions In Click-Through Rate Prediction

Ze Meng, Jinnian Zhang, Yumeng Li, Jiancheng Li, Tanchao Zhu, Lifeng Sun, A General Method For Automatic Discovery of Powerful Interactions In Click-Through Rate Prediction. ACM Special Interest Group on Information Retrieval (SIGIR), 2021.
Code

MMFashion: An Open-Source Toolbox for Visual Fashion Analysis

Xin Liu, Jiancheng Li, Jiaqi Wang, Ziwei Liu, MMFashion: An Open-Source Toolbox for Visual Fashion Analysis. ACM Multimedia (ACM MM), 2021.
Code

TreeNet: Learning Sentence Representations with Unconstrained Tree Structure

Zhou Cheng, Chun Yuan, Jiancheng Li, Haiqin Yang, TreeNet: Learning Sentence Representations with Unconstrained Tree Structure. International Joint Conference on Artificial Intelligence (IJCAI), 2018.
Code

Gate Function Based Structure Adaptive Convolution for Scene Semantic Segmentation

Zhou Cheng, Jiancheng Li, Chun Yuan, Gate Function Based Structure Adaptive Convolution for Scene Semantic Segmentation. IEEE International Conference on Multimedia and Expo (ICME), 2017.

Automatic Image Annotation Using Adaptive Weighted Distance in Improved K Nearest Neighbors Framework

Jiancheng Li, Chun Yuan, Automatic Image Annotation Using Adaptive Weighted Distance in Improved K Nearest Neighbors Framework, IEEE Pacific Rim Conference on Multimedia. Springer International Publishing (PCM), 2016.

Multi-label Image Annotation Based on Convolutional Neural Network

Jiancheng Li, Chun Yuan, You Song, Multi-label Image Annotation Based on Convolutional Neural Network, Computer Science, P.R.C, 2016.

Qichedashi Text Summarization

Given multiple rounds of dialogue between the technician and the user, output a diagnostic advice report. Main work: Encoder/Decoder: stacked bi-directional LSTM, attention, pre-processing (stop words, etc.), post-processing (matching the original sentence, etc.).
Team: Jiancheng Li

2018.05.11 - 2018.06.28 PaddlePaddle AI Industry Application Competition Qichedashi Text Summarization Rank 3

DuReader Machine Reading Comprehension

Given questions and related documents, output the answers. Main work: Add self-attention, use Chinese word vector word2vec-CBOW-cn, and train a paragraph ranking model.
Team: Jiancheng Li

2018.04.26 - 2018.06.25 Baidu PaddlePaddle AI Competition Machine Reading Comprehension Rank 1

Truefan-X

A VR app that plays 360 degree streamed live video in android with VR devices such as Google Cardboard / Daydream. The app is mainly to watch live football games using VR devices which makes the user feel like in the stadium. My duty is to build the android app that can play 360 degree streamed live video using VR devices.
Team: Simon Pfältzer, Julian Zachmann, Manuel Geitner, Jiancheng Li, Connor Lurring

2018.01.19 - 2018.01.22 FCBayernHackDays | Page | Code

Classification and Detection on X-ray images

Given frontal-view X-ray images, implement a multi-label image classification model to recognize the pathology and an object detection model to localize the pathology.
Team: Jiancheng Li, Yang Gu

2017.12.16 IBM Cognitive Systems PowerAI hackathon finals first prize

wish list

A facebook app about wish list: add an "add to my fb wish list" button in the shopping website, show and search my friends' and my wish list in the facebook app. My duty is using the facebook api and writing the backend code.
Team: Jeremy Fu, Jiancheng Li, Kimmers Sing, Chris Le

2017.04.30 Facebook Sydney Hackathon 2017 | Demo | Code

HWCC2017

A tool that solves a facility location problem.
Team: Jiancheng Li

2017.04 2017 Huawei CodeCraft - Regional Silver Medal (round 2 Top 36) | Code

scikit-learn-doc-zh

scikit-learn doc Chinese translation. Provides docker image.
Team: Jiancheng Li

2016.02 - now Open Source Project | Doc | Code

RayTracer

Ray Tracing demos in Windows Form.
Team: Jiancheng Li

2016.01 Project of course "Computer Graphics" | Code

MicroCast

A cooperative video streaming tool on Raspberry Pi.
Team: Jiancheng Li, Yuntao Jiang

2016.01 Project of course "Video Streaming" | Code

Poem Generator

An online Chinese poem generator that generates poems according to the corpus of Tang poems by denoising preprocessing, word segmentation, collocation generation and theme generation.
Team: Jiancheng Li

2015.12 Project of course "Computational Linguistics" | Blog | Code

Tag Extraction of Chinese Text

Tag extraction of Chinese Text using IBM Linux POWER 8. Using Natural Language Processing (NLP) methods.
Team: Jiancheng Li, Weibin Lv

2015.12.12 IBM Linux on Power Algorithm Marathon Challenge 3rd Prize

AudioCity

An android app that provides audio information of places, such as museums, libraries and so on. Competed in InnoApps Hackathon 2014 which the competition theme is Smart City.
Team: Jiancheng Li, Andrei Diaconu

2015.02.04 InnoApps Hackathon 2014 (European Young Innovators Forum) 3rd Place

Plask

An online questionnaire website using Python and Flask. Rapidly developed in two weeks and follows the agile development method. Team: Jiancheng Li, Ruoyu Fan, Siyu Yang, Hanxiong Wang, Xiangyu Ji, Junyi Du, Qi Liu, Yang Yang, Jiachen Chen

2014.06 - 2014.07 Flask version: Project in the 3rd semester. Agile development. | Code

Let's Date

A website about the management of date, meeting and other activities. Based on PHP.
Team: Jiancheng Li

2014.03 2013 Open Source Application and Innovation Challenge | Code

TestSame

A tool that calculates code similarity based on .NET Framework.
Team: Jiancheng Li

2013.08 A tool when I was the teaching assistant of course "High-level language programming". | Code

UChange

A website about second-hand goods exchange.
Team: Jiancheng Li, Ruoyu Fan, Siyu Yang, Xiangyu Ji, Hanxiong Wang

2013.03 - 2013.06 Django version: Project of course "IT Innovation and Business", Project of course "Software Engineering", Student Research Training Project (SRTP)
2013.08 - 2013.09 ASP.NET version: Project in the 3rd semester. | Code
2014.06 SSH version: Project of course "J2EE" | Code

HandyBUAA

An iOS app that provides information of Beihang campus.
Team: Yiling Chen, Jiancheng Li

2012.11 - 2013.04 App store

Ocean Scream

An AI game platform that two AI players can control their own ships to move and battle and the ships can be recovered in some islands. My duty is the AI SDK and network connection programming, the SDK supports C#, C++ and Java.
Team: BUAA-MSTC (Microsoft Student Technology Club of Beihang University)

2012.03 - 2012.05 2012 Microsoft Beauty of Programming Competition Platform | Video | Code

Unreal Museum

A VR demo that shows virtual historical relics in virtual museum. It uses a Kinect to detect the position of human head, and apply a calculated matrix between the view matrix and the projection matrix (or, alter the projection matrix) of a 3D scene. Thus creates an illusion of the reality for the screen no longer appears as an 2D plane when you see the screen from different directions. My duty is project management and programming.
Team: Jiancheng Li, Ruoyu Fan | Video | Code

2011.11 - 2012.04 2012 Microsoft Elite Challenge (Microsoft Research Asia) 3rd Prize (Team name: Funkect)
2012.01 - 2012.05 Microsoft Imagine Cup 2012 - passed round 1
2012.04 The 22nd Fengru Cup of Beihang University 2nd Prize
2013.06 The 1st Kinect Geek Innovation Application Competition Honorable Mention

Dango's Adventure

An ACT game that moves the cloud at the bottom of the screen to collect the dangos that fell from the top.
Team: Jiancheng Li

2011.11 XNA
2012.02 Java
2012.04 Flash
2012.06 Android
2013.09 The 4th Android Application Development Student Challenge (Beijing region) Honorable Mention
2012.08 iOS
2012.10 Windows Phone
2013.03 HTML5 | Play! | Code

A General Method For Automatic Discovery of Powerful Interactions In Click-Through Rate Prediction

Ze Meng, Jinnian Zhang, Yumeng Li, Jiancheng Li, Tanchao Zhu, Lifeng Sun, A General Method For Automatic Discovery of Powerful Interactions In Click-Through Rate Prediction. ACM Special Interest Group on Information Retrieval (SIGIR), 2021.
代码

MMFashion: An Open-Source Toolbox for Visual Fashion Analysis

Xin Liu, Jiancheng Li, Jiaqi Wang, Ziwei Liu, MMFashion: An Open-Source Toolbox for Visual Fashion Analysis. ACM Multimedia (ACM MM), 2021.
代码

TreeNet: Learning Sentence Representations with Unconstrained Tree Structure

Zhou Cheng, Chun Yuan, Jiancheng Li, Haiqin Yang, TreeNet: Learning Sentence Representations with Unconstrained Tree Structure. International Joint Conference on Artificial Intelligence (IJCAI), 2018.
代码

Gate Function Based Structure Adaptive Convolution for Scene Semantic Segmentation

Zhou Cheng, Jiancheng Li, Chun Yuan, Gate Function Based Structure Adaptive Convolution for Scene Semantic Segmentation. IEEE International Conference on Multimedia and Expo (ICME), 2017.

Automatic Image Annotation Using Adaptive Weighted Distance in Improved K Nearest Neighbors Framework

Jiancheng Li, Chun Yuan, Automatic Image Annotation Using Adaptive Weighted Distance in Improved K Nearest Neighbors Framework, IEEE Pacific Rim Conference on Multimedia. Springer International Publishing (PCM), 2016.

基于卷积神经网络的多标签图像自动标注

黎健成, 袁春, 宋友, 基于卷积神经网络的多标签图像自动标注, 计算机科学, 2016.

汽车大师文本摘要

给定技师与用户的多轮对话，输出诊断建议报告。主要工作：Encoder/Decoder: stacked bi-directional LSTM, attention，预处理（停用词等）、后处理（匹配原句等）。
团队：Jiancheng Li

2018.05.11 - 2018.06.28 PaddlePaddle AI 产业应用赛——汽车大师问答摘要与推理 Rank 3