Zobrazeno 1 - 10
of 53
pro vyhledávání: '"Li-po Han"'
Multimodal encoders like CLIP excel in tasks such as zero-shot image classification and cross-modal retrieval. However, they require excessive training data. We propose canonical similarity analysis (CSA), which uses two unimodal encoders to replicat
Externí odkaz:
http://arxiv.org/abs/2410.07610
Image retrieval is crucial in robotics and computer vision, with downstream applications in robot place recognition and vision-based product recommendations. Modern retrieval systems face two key challenges: scalability and efficiency. State-of-the-a
Externí odkaz:
http://arxiv.org/abs/2410.07022
Autor:
Narayanan, Aditya, Kasibhatla, Pranav, Choi, Minkyu, Li, Po-han, Zhao, Ruihan, Chinchali, Sandeep
Networked robotic systems balance compute, power, and latency constraints in applications such as self-driving vehicles, drone swarms, and teleoperated surgery. A core problem in this domain is deciding when to offload a computationally expensive tas
Externí odkaz:
http://arxiv.org/abs/2409.06078
Autor:
Li-Po Han, Cheng-Yue Zhang, Feng-Xian Wang, Li-Li Zang, Dong-Yue Liu, Qiao-Zhi Ma, Wei-Zhong Wang, Lan Qi
Publikováno v:
Guoji Yanke Zazhi, Vol 21, Iss 9, Pp 1641-1643 (2021)
AIM: To analyze the anatomical characteristics of nasolacrimal duct development in children with congenital nasolacrimal duct obstruction by nasolacrimal duct CT scan combined with three-dimensional reconstruction technology.METHODS: Prospective case
Externí odkaz:
https://doaj.org/article/08f4ce163c73469eaf37380dfc1cfc1c
Publikováno v:
Guoji Yanke Zazhi, Vol 20, Iss 9, Pp 1649-1652 (2020)
AIM: To observe and analyze the therapeutic effect of fine training combined with virtual reality brain vision training in amblyopic children.METHODS: Case control study. 232 cases(416 eyes)of amblyopia were diagnosed in Baoding children's Hospital f
Externí odkaz:
https://doaj.org/article/465816248536472d80cab50b19154ffd
Foundation models have recently expanded into robotics after excelling in computer vision and natural language processing. The models are accessible in two ways: open-source or paid, closed-source options. Users with access to both face a problem whe
Externí odkaz:
http://arxiv.org/abs/2402.08570
Autor:
Li, Po-han, Ankireddy, Sravan Kumar, Zhao, Ruihan, Mahjoub, Hossein Nourkhiz, Moradi-Pari, Ehsan, Topcu, Ufuk, Chinchali, Sandeep, Kim, Hyeji
Publikováno v:
NeurIPS 2023
Efficient compression of correlated data is essential to minimize communication overload in multi-sensor networks. In such networks, each sensor independently compresses the data and transmits them to a central node due to limited communication bandw
Externí odkaz:
http://arxiv.org/abs/2305.15523
Fleets of networked autonomous vehicles (AVs) collect terabytes of sensory data, which is often transmitted to central servers (the ''cloud'') for training machine learning (ML) models. Ideally, these fleets should upload all their data, especially f
Externí odkaz:
http://arxiv.org/abs/2303.03602
We analyze a cost-minimization problem in which the controller relies on an imperfect timeseries forecast. Forecasting models generate imperfect forecasts because they use anonymization noise to protect input data privacy. However, this noise increas
Externí odkaz:
http://arxiv.org/abs/2210.00358
We propose a method to attack controllers that rely on external timeseries forecasts as task parameters. An adversary can manipulate the costs, states, and actions of the controllers by forging the timeseries, in this case perturbing the real timeser
Externí odkaz:
http://arxiv.org/abs/2207.06982