Short Links

Goal

Some students complained that university (especially PKU and THU, or even MIT) is overrated:

• SCHOOL = Seven Crappy Hours Of Our Life
• CLASS = Come Late And Start Sleeping
• FINALS = F**k, I Never Actually Learned S**t

I started to build a list for students who hope to join a modern research lab in AI. This list should give you a relatively good starting point to learn at your own pace.

Motivation

Profs. Yixin Zhu and Yizhou Wang’s groups are developing computer vision algorithms to detect and identify primates (monkeys and apes) and their behaviors. They are working with primate video data from their collaborators of Professor Federico Rossano’s Comparative Cognition Laboratory (CCL) at the University of California, San Diego (UCSD). One of their projects uses video footage of motion-triggered trail cameras that are installed in the enclosures of several different groups of macaque monkeys that are housed at a large animal sanctuary in the United States. Due to large enclosure sizes at this primate sanctuary, the animal caretakers spend significant amounts of their time on monitoring the well-being of their animals. The goal of this project is to build a system for the caretakers that uses the above-mentioned computer vision algorithms to detect and identify specific macaques from trail camera footage. This will help the animal sanctuary to provide better care for their animals and can further aid behavioral researchers at CCL. The challenge and could be scaled to other sanctuaries in the future.

Motivation

This research project entails collaborating with scientists at the Comparative Cognition Laboratory (CCL) at the University of California, San Diego (UCSD). Comparative Psychologists study the evolutionary origins of human behavior and cognition by investigating similarities and differences between humans and other animals. They commonly employ mechanical apparatuses to test the cognitive abilities of nonhuman animals, such as nonhuman primates, in controlled experimental settings.

Background

Understanding the behavior of primates is important for primatology, psychology, and for biology more broadly since primates are an important model organism, and whose behavior is often an important variable of interest. However, our ability to rigorously quantify behavior has long been limited, due to the high manual cost of observing and the difficulty in long-term observing in the wild. Behavioral tracking in primates has recently become possible through parallel developments in computer vision, machine learning, and robotics when technical breakthroughs in deep learning enabled software to recognize objects in an image by use of CNN. Several works enable automatic tracking of humans (Cao et al., 2019 Newell et al., 2016). These works in turn inspired and facilitated work that allowed for the tracking of animals such as flies, mice, and horses from videos (Mathis et al., 2018; Pereira et al., 2019).

Background

A receiver function (RF) is the response of the Earth structure below a seismometer to an incident seismic wave and consists of a series of P-to-S (Ps) or S-to-P (Sp) converted waves generated at structural interfaces, mainly those generated by the velocity discontinuities in the crust and upper mantle below the seismometer. Crustal thickness (H) and P-wave and S-wave velocity ratio (κ) are important parameters to control the RFs features. These two parameters can reflect the crustal structure and internal material composition. They can provide an essential basis for regional tectonics and dynamics. Here, we want to develop a pattern recognition method to extract the H and κ from RFs. In a way, this is an inversion problem.

Background

Global Seismographic Network (GSN) is deployed worldwide to listen to the sound of the Earth. The recorded signals, also referred to as seismograms, contains rich information on the environmental processes, seismic hazard, and could also be used to monitor human activities.