Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (4 results)

  • Thesis
  • Peer Reviewed
UC Riverside Electronic Theses and Dissertations (2014)

In this work, we focus on the problem of pose estimation in unknown environments, using the measurements from an inertial measurement unit (IMU) and a single camera. We term this estimation task visual-inertial odometry (VIO), in analogy to the well-known visual-odometry (VO) problem. Our focus is on developing VIO algorithms for high-precision, consistent motion estimation in real time. The majority of VIO algorithms proposed to date have been developed for systems which are equipped with high-end processors and high-quality sensors. By contrast, we are interested in tracking the motion of systems that are small and inexpensive, and are equipped with limited processing and sensing resources.

Such resource-constrained systems are common in application areas such as micro aerial vehicles, mobile phones, and augmented reality (AR) devices. Endowing such systems with the capability to accurately track their poses will create a host of new opportunities for commercial applications, and lower the barrier to entry in robotics research and development.

Performing accurate motion estimation on resource-constrained systems requires novel methodologies to address the challenges caused by the limited sensing and processing capabilities, and to provide guarantees for the optimal utilization of these resources. To this end, in this work, we focus on developing novel, resource-

adaptive VIO algorithms based on the extended Kalman filter (EKF) formulation. Specifically, we (i) analyze the properties and performance of existing EKF-based

VIO approaches, and propose a novel estimator design method, which ensures the correct observability properties of the linearized system models to improve the estimates' accuracy and consistency, (ii) present a methodology for minimizing the computational cost of the EKF-VIO algorithms, which relies on online optimization of the estimator's parameters based on the properties of the environment, (iii) propose an algorithm for joint online calibration of the spatial and temporal relationship between the visual and inertial sensors, and (iv) propose high-fidelity sensor models that enable us to process the measurements captured by rolling-shutter cameras and low-cost inertial sensors. We evaluate our estimators with various simulated and real-world data sets, which demonstrate that our proposed formulations are able to consistently and accurately track the pose of resource-constrained systems in real time.

    Cover page: Visual-Inertial Odometry on Resource-Constrained Systems
    • Article
    • Peer Reviewed
    UC Riverside Previously Published Works (2017)
      Cover page: A dense flow-based framework for real-time object registration under compound motion
      • Article
      • Peer Reviewed
      UC Santa Cruz Previously Published Works (2019)
      Ultrafast-charging energy storage devices are attractive for powering personal electronics and electric vehicles. Most ultrafast-charging devices are made of carbonaceous materials such as chemically converted graphene and carbon nanotubes. Yet, their relatively low electrical conductivity may restrict their performance at ultrahigh charging rate. Here, we report the fabrication of a porous titanium nitride (TiN) paper as an alternative electrode material for ultrafast-charging devices. The TiN paper shows an excellent conductivity of 3.67 × 104 S m-1, which is considerably higher than most carbon-based electrodes. The paper-like structure also contains a combination of large pores between interconnected nanobelts and mesopores within the nanobelts. This unique electrode enables fast charging by simultaneously providing efficient ion diffusion and electron transport. The supercapacitors (SCs) made of TiN paper enable charging/discharging at an ultrahigh scan rate of 100 V s-1 in a wide voltage window of 1.5 V in Na2SO4 neutral electrolyte. It has an outstanding response time with a characteristic time constant of 4 ms. Significantly, the TiN paper-based SCs also show zero capacitance loss after 200,000 cycles, which is much better than the stability performance reported for other metal nitride SCs. Furthermore, the device shows great promise in scalability. The filtration method enables good control of the thickness and mass loading of TiN electrodes and devices.
        Cover page: TiN Paper for Ultrafast-Charging Supercapacitors.
        • Article
        • Peer Reviewed
        UC Berkeley Previously Published Works (2022)
        This work proposes a new computational framework for learning a structured generative model for real-world datasets. In particular, we propose to learn a Closed-loop Transcriptionbetween a multi-class, multi-dimensional data distribution and a Linear discriminative representation (CTRL) in the feature space that consists of multiple independent multi-dimensional linear subspaces. In particular, we argue that the optimal encoding and decoding mappings sought can be formulated as a two-player minimax game between the encoder and decoderfor the learned representation. A natural utility function for this game is the so-called rate reduction, a simple information-theoretic measure for distances between mixtures of subspace-like Gaussians in the feature space. Our formulation draws inspiration from closed-loop error feedback from control systems and avoids expensive evaluating and minimizing of approximated distances between arbitrary distributions in either the data space or the feature space. To a large extent, this new formulation unifies the concepts and benefits of Auto-Encoding and GAN and naturally extends them to the settings of learning a both discriminative and generative representation for multi-class and multi-dimensional real-world data. Our extensive experiments on many benchmark imagery datasets demonstrate tremendous potential of this new closed-loop formulation: under fair comparison, visual quality of the learned decoder and classification performance of the encoder is competitive and arguably better than existing methods based on GAN, VAE, or a combination of both. Unlike existing generative models, the so-learned features of the multiple classes are structured instead of hidden: different classes are explicitly mapped onto corresponding independent principal subspaces in the feature space, and diverse visual attributes within each class are modeled by the independent principal components within each subspace.
          Cover page: CTRL: Closed-Loop Transcription to an LDR via Minimaxing Rate Reduction
          Top