Source author record

Hongbo Zhang

Hongbo Zhang appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Machine Learning Robotics Artificial Intelligence astro-ph.CO Computation and Language Computer Vision

Catalog footprint

What is connected

5works

6topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

FAAST: Forward-Only Associative Learning via Closed-Form Fast Weights for Test-Time Supervised Adaptation

Adapting pretrained models typically involves a trade-off between the high training costs of backpropagation and the heavy inference overhead of memory-based or in-context learning. We propose FAAST, a forward-only associative adaptation method that analytically compiles labeled examples into fast weights in a single pass. By eliminating memory or context dependence, FAAST achieves constant-time inference and decouples task adaptation from pretrained representation. Across image classification and language modeling benchmarks, FAAST matches or exceeds backprop-based adaptation while reducing adaptation time by over 90% and is competitive to memory/context-based adaptation while saving memory usage by up to 95%. These results demonstrate FAAST as a highly efficient, scalable solution for supervised task adaptation, particularly for resource-constrained models. We release the code and models at https://github.com/baoguangsheng/faast.

preprint2022arXiv

GenLoco: Generalized Locomotion Controllers for Quadrupedal Robots

Recent years have seen a surge in commercially-available and affordable quadrupedal robots, with many of these platforms being actively used in research and industry. As the availability of legged robots grows, so does the need for controllers that enable these robots to perform useful skills. However, most learning-based frameworks for controller development focus on training robot-specific controllers, a process that needs to be repeated for every new robot. In this work, we introduce a framework for training generalized locomotion (GenLoco) controllers for quadrupedal robots. Our framework synthesizes general-purpose locomotion controllers that can be deployed on a large variety of quadrupedal robots with similar morphologies. We present a simple but effective morphology randomization method that procedurally generates a diverse set of simulated robots for training. We show that by training a controller on this large set of simulated robots, our models acquire more general control strategies that can be directly transferred to novel simulated and real-world robots with diverse morphologies, which were not observed during training.

preprint2021arXiv

Open-set Intersection Intention Prediction for Autonomous Driving

Intention prediction is a crucial task for Autonomous Driving (AD). Due to the variety of size and layout of intersections, it is challenging to predict intention of human driver at different intersections, especially unseen and irregular intersections. In this paper, we formulate the prediction of intention at intersections as an open-set prediction problem that requires context specific matching of the target vehicle state and the diverse intersection configurations that are in principle unbounded. We capture map-centric features that correspond to intersection structures under a spatial-temporal graph representation, and use two MAAMs (mutually auxiliary attention module) that cover respectively lane-level and exitlevel intentions to predict a target that best matches intersection elements in map-centric feature space. Under our model, attention scores estimate the probability distribution of the openset intentions that are contextually defined by the structure of the current intersection. The proposed model is trained and evaluated on simulated dataset. Furthermore, the model, trained on simulated dataset and without any fine tuning, is directly validated on in-house real-world dataset collected at 98 realworld intersections and exhibits satisfactory performance,demonstrating the practical viability of our approach.

preprint2020arXiv

Triple-GAIL: A Multi-Modal Imitation Learning Framework with Generative Adversarial Nets

Generative adversarial imitation learning (GAIL) has shown promising results by taking advantage of generative adversarial nets, especially in the field of robot learning. However, the requirement of isolated single modal demonstrations limits the scalability of the approach to real world scenarios such as autonomous vehicles' demand for a proper understanding of human drivers' behavior. In this paper, we propose a novel multi-modal GAIL framework, named Triple-GAIL, that is able to learn skill selection and imitation jointly from both expert demonstrations and continuously generated experiences with data augmentation purpose by introducing an auxiliary skill selector. We provide theoretical guarantees on the convergence to optima for both of the generator and the selector respectively. Experiments on real driver trajectories and real-time strategy game datasets demonstrate that Triple-GAIL can better fit multi-modal behaviors close to the demonstrators and outperforms state-of-the-art methods.

preprint2011arXiv

VLBI observations of 10 CSO candidates: expansion velocities of hot spots

Observations of ten Compact Symmetric Objects ({\rm CSO}) candidates have been made with the Very Long Baseline Array at 8.4 GHz in 2005 and with a combined Chinese and European VLBI array at 8.4 GHz in 2009. The 2009 observations incorporate for the first time the two new Chinese telescopes at Miyun and Kunming for international astrophysical observations. The observational data, in combination with archival VLBA data from previous epochs, have been used to derive the proper motions of the VLBI components. Because of the long time baseline of $\sim$16 years of the VLBI data sets, the expansion velocities of the hot spots can be measured at an accuracy as high as $\sim$1.3 $μ$as yr$^{-1}$. Six of the ten sources are identified as CSOs with a typical double or triple morphology on the basis of both spectral index maps and their mirror-symmetry of proper motions of the terminal hot spots. The compact double source J1324+4048 is also identified as a CSO candidate. Among the three remaining sources, J1756+5748 and J2312+3847 are identified as core-jet sources with proper motions of their jet components relating to systemic source expansion. The third source J0017+5312 is likely also a core-jet source, but a robust detection of a core is needed for an unambiguous identification. The kinematic ages of the CSOs derived from proper motions range from 300 to 2500 years. The kinematic age distribution of the CSOs confirm an overabundance of compact young CSOs with ages less than 500 years. CSOs with known kinematic ages may be used to study the dynamical evolution of extragalactic radio sources at early stages.