CS/EE student at KAIST. Winner of 10+ Physics Olympiads including International Physics Olympiad 2 times. Strong expertise in Data Structures and Algorithms. Specialization in ML and Deep Learning.
ประวัติการทำงาน
โพสต์
สรุปประวัติการทำงานโดย AI
술레이만자데아이한님은 KAIST에서 CS/EE를 전공하며 기하학적 딥러닝 및 응용 분야 연구 인턴으로 활동했습니다. 에버엑스에서는 3D 인간 자세 추정 및 그래프 신경망 연구를 수행하며 SOTA(State-of-the-Art) 성능을 달성한 경험이 있습니다. 또한, 올림피아드 코치 및 팀 리더로서 다수의 학생들을 지도하여 메달 획득을 이끈 경험이 있습니다.
• Researched on 3D Human Pose Estimation task and Graph Neural Networks. Analyzed and replicated related works Gong et al. and Choi et al.
• Proposed the novel model and based on Diffusion and Graph Convolutional Network (GCN) models for Human3.6M and HumanEva‐I datasets. Designed and applied temporal convolutional network and temporal transformers for handling information between video frames.
• Obtained average MPJPE error of 8.93 mm and 12.22 mm for Walking and Jogging actions in HumanEva‐I dataset, which were the SOTA results by improvement of 1.4 mm and 0.9 mm respectively.
• Worked on integration of better object detection models such as YOLOv7 to KAPAO model on company dataset.
Tutored and mentored freshman students in Introduction to Programming from March to June 2022 on a weekly basis.
Tutoring and mentoring freshman students in Calculus II on a weekly basis starting from September 2022
Ministry of Science and Education Republic of Azerbaijan
2021년 9월 - 2022년 9월 · 1년 1개월
• Represented Azerbaijan in 52nd International Physics Olympiad as the Team Leader, ensured the detailed and accurate translations of problems
and attend in Internatonal Board Meeting and coordination process.
• Proposed new and original problems for the national team members, as well as for team selection exams in
physics as a member of Scientific Committee . Actively participated in problem selection, grading, and the coordination process of International Young Talents Olympiad. Graded solutions of more than 100 students from 15 countries.
• Taught Physics Olympiad lessons to junior students who are preparing for Physics Olympiads as Olympiad Coach. Assisted students with their tasks and
other learning difficulties. Kept track of learner performance by conduction exams on totally original problems. Mentored students to get 4
bronze medals in European Physics Olympiad and 1 bronze medal in International Physics Olympiad
The special award to the top 3% of the students from each department ranked by GPA
프로젝트
Semi‑Supervised Semantic Segmentation with Cross‑Consistency Training
2022년 9월 - 2022년 12월 · 4개월
Implemented SSSS using 1 shared encoder and several (main and 40 auxilirary) decoder models. Augmented the base model with Temporal Ensembling model, GAN, and encoder architecture changes. Managed to improve the results of core implementation from mIoU score of 73.2 to 73.5 using Seg-GAN network.
프로젝트
Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
2022년 9월 - 2022년 12월 · 4개월
Replicated the bottom-up top-down attention for Image Captioning. Augmented the base model with an Auto-Reconstructor Network, an Adaptive Attention mechanism via a Visual Sentinel, and our novel approach of salient attention via salience map features. Managed to obtain comparable good results.
프로젝트
UNIX Shell Implementation
2022년 9월 - 2022년 12월 · 4개월
Implemented the program that simulates command line interface in Unix environment and it consists of a lexical analyzer and a syntactic analyzer. Managed to make program to handle both built‑in/external commands, redirection operations, signals and pipes with no memory leak.
