A5058375642- Planetary Science and Exploration
- Domain Adaptation and Few-Shot Learning
- Anomaly Detection Techniques and Applications
- Advanced Memory and Neural Computing
- Ferroelectric and Negative Capacitance Devices
- Artificial Immune Systems Applications
- Data Stream Mining Techniques
- Space Exploration and Technology
- Spaceflight effects on biology
- Astro and Planetary Science
- Marine and environmental studies
- Colorectal Cancer Screening and Detection
- Transition Metal Oxide Nanomaterials
- Geophysics and Gravity Measurements
- Methane Hydrates and Related Phenomena
- Geophysical Methods and Applications
- Advanced Neural Network Applications
- Space Science and Extraterrestrial Life
- Multimodal Machine Learning Applications
- Network Security and Intrusion Detection
- AI in cancer detection
- Radiomics and Machine Learning in Medical Imaging
- Human Pose and Action Recognition
Tongji University
2024-2025
Arizona State University
2021-2022
The identification of optimal landing sites is a critical first step for successful missions to the Moon and other extraterrestrial bodies, necessitating integration various environmental factors over large spatial scales. At lunar south pole, site selection must balance engineering safety with areas high scientific interest, requiring extensive analysis potential locations. Although intelligent algorithms have been increasingly investigated this purpose, application deep learning techniques...
Resistive random-access memory (RRAM)-based in-memory computing (IMC) architectures offer an energy-efficient solution for DNN acceleration. However, the performance of RRAM-based IMC is limited by device nonidealities, ADC precision, and algorithm properties. To address this, in this work, first, we perform statistical characterization RRAM variation temporal degradation from 300mm wafers a fully integrated CMOS/RRAM 1T1R test chip at 65nm. Through build realistic foundation to assess...
RRAM-based in-memory computing (IMC) effectively accelerates deep neural networks (DNNs). Furthermore, model compression techniques, such as quantization and pruning, are necessary to improve algorithm mapping hardware performance. However, in the presence of RRAM device variations, low-precision sparse DNNs suffer from severe post-mapping accuracy loss. To address this, this work, we investigate a new metric, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Novelty detection aims to automatically identify out-of-distribution (OOD) data, without any prior knowledge of them. It is a critical step in data monitoring, behavior analysis and other applications, helping enable continual learning the field. Conventional methods OOD perform multi-variate on an ensemble or features, usually resort supervision with improve accuracy. In reality, such impractical as one cannot anticipate anomalous data. this paper, we propose novel, self-supervised approach...
The quest for water-ice is critical in the scientific exploration of Mars, as areas where water reservoirs have been identified could be attractive landing sites future missions. Korolev crater located at Martian north pole, and interior known its distinct layer bright deposits exposed surface, probably composed rocks. Although studies estimated depth deposits, effective volume ice remains undisclosed. We quantified within using SHARAD data, supported by global MOLA DEM Context Camera...
Continual learning, the capability to learn new knowledge from streaming data without forgetting previous knowledge, is a critical requirement for dynamic learning systems, especially emerging edge devices such as self-driving cars and drones. However, continual still facing catastrophic problem. Previous work illustrate that model performance on not only related algorithms but also strongly dependent inherited model, i.e., where starts. The better stability of less thus, should be...