Corrina Wu Email & Phone Number

Mountain View, California, United States

- Played a key role in solving complex technical problems, particularly in creating algorithms for optimal prompt pixel selection for SAM, leading to increased precision and operational speed- Pioneered the development and execution of advanced preprocessing techniques for leaf image analysis, crucial for a deep learning project focused on tree species recognition- Innovated in the design of algorithms for prompt engineering, markedly improving the efficiency of binary mask generation from leaf images using the Segment Anything Model (SAM), achieving a 50% boost in efficiency- Tackled the challenge of a limited dataset by the training data enrichment process, enhancing the model’s learning capabilities and accuracy in species identification

Berkeley, California, United States

- Collaborated with a cross-functional Agile team to design and launch a food exploration app featuring multimedia content on both the App Store and Google Play- Designed and engineered the application's backend foundation, leveraging AWS technologies such as EC2, S3, Cognito, and CDN to focus on scalability, system reliability, and robust security- Developed 10+ robust APIs to facilitate core functionalities including video posting, user authentication, as well as video commenting and liking, utilizing Node.js and MongoDB- Optimized video upload API, reducing AWS S3 storage expenses by 20% and improving upload speed by 20%- Crafted an intuitive frontend using React Native, incorporating 10+ versatile components to streamline navigation, layout, and user actions, bolstering overall app usability

- Created a user-friendly web portal for UPE to streamline access to various services for UC Berkeley students, tailored to their UPE status (candidates/members/other students)- Skillfully designed and developed the portal's frontend using Figma and React, enhancing information accessibility and improving user experience for the UC Berkeley community

Berkeley, California, United States

- Collaborated with SETI team to create an automated signal analysis pipeline for extraterrestrial intelligence detection, handling petabytes of global telescope data- Implemented ResNet50 in Python for optimized feature extraction from raw telescope signal data- Utilized reverse image searches and sklearn's nearest neighbors for interference filtering and pattern recognition- Transformed manual signal analysis process into an automated program, reducing analysis time by 80%- Advanced the search for potential extraterrestrial signals through efficient processing of massive radio telescope data

Zhuhai, Guangdong, China

- Developed an Android authentication platform to streamline user access across thousands of applications and games using Android Studio and Java- Integrated over 40 external account options, including Google and WeChat, to provide seamless authentication for millions of users- Fortified security by leveraging bcrypt for password hashing and implemented JWT for stateless authentication in RESTful API- Achieved a 60% reduction in user account registration time, optimizing the user registration process and enhancing user experience- Collaborated with diverse teams to ensure efficient and timely delivery of essential features and functionalities for project success

Berkeley, California, United States

- Developed a Python program to automate sulci matching in brain data, utilizing parameters such as vertex coordinates, thickness, and depth of curves, resulting in an 80% increase in efficiency- Identified and labeled active fMRI regions for 50+ subjects, facilitating cortical mapping of natural emotional faces and enabling comparisons between subject groups- Recognized for outstanding contributions to the "Multi-feature Modeling of Cortical Tuning to Natural Emotional Images" project, showcasing a combination of cognitive neuroscience and programming expertise- Applauded by supervisor for collaboration and coordination with peers, actively contributing to cutting-edge research in cognitive neuroscience while demonstrating strong teamwork abilities

Zhuhai, Guangdong, China

- Analyzed GPS data to predict vehicle traffic flow in Zhuhai, Guangdong (a city with about 2.5 million people), aiding traffic police and map platforms, potentially reducing travel time by 15% during peak hours- Utilized sklearn models to analyze airline traffic and optimize flight ticket pricing, potentially increasing profits by 20% for ticket platforms- Processed geospatial company data within polygon networks to identify cost-effective locations for 5G tower installation, resulting in 20% cost savings- Designed compelling visual aids and data visualizations, using tools such as Matplotlib and Seaborn, to effectively communicate insights