skip to main content


Search for: All records

Creators/Authors contains: "Zhang, Qian"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Depth estimation is fundamental to 3D perception, and humans are known to have biased estimates of depth. This study investigates whether convolutional neural networks (CNNs) can be biased when predicting the sign of curvature and depth of surfaces of textured surfaces under different viewing conditions (field of view) and surface parameters (slant and texture irregularity). This hypothesis is drawn from the idea that texture gradients described by local neighborhoods—a cue identified in human vision literature—are also representable within convolutional neural networks. To this end, we trained both unsupervised and supervised CNN models on the renderings of slanted surfaces with random Polka dot patterns and analyzed their internal latent representations. The results show that the unsupervised models have similar prediction biases as humans across all experiments, while supervised CNN models do not exhibit similar biases. The latent spaces of the unsupervised models can be linearly separated into axes representing field of view and optical slant. For supervised models, this ability varies substantially with model architecture and the kind of supervision (continuous slant vs. sign of slant). Even though this study says nothing of any shared mechanism, these findings suggest that unsupervised CNN models can share similar predictions to the human visual system. Code: github.com/brownvc/Slant-CNN-Biases 
    more » « less
    Free, publicly-accessible full text available August 5, 2024
  2. Parallel tempering (PT), also known as replica exchange, is the go-to workhorse for simulations of multi-modal distributions. The key to the success of PT is to adopt efficient swap schemes. The popular deterministic even-odd (DEO) scheme exploits the non-reversibility property and has successfully reduced the communication cost from O(P 2) to O(P) given sufficient many P chains. However, such an innovation largely disappears in big data problems due to the limited chains and extremely few bias-corrected swaps. To handle this issue, we generalize the DEO scheme to promote the non-reversibility and obtain an appealing communication cost O(P log P) based on the optimal window size. In addition, we also analyze the bias when we adopt stochastic gradient descent (SGD) with large and constant learning rates as exploration kernels. Such a user-friendly nature enables us to conduct large-scale uncertainty approximation tasks without much tuning costs. 
    more » « less
  3. Abstract Eukaryotic cells have evolved membrane-bound organelles, including the endoplasmic reticulum (ER), Golgi, mitochondria, peroxisomes, chloroplasts (in plants and green algae) and lysosomes/vacuoles, for specialized functions. Organelle quality control and their proper interactions are crucial both for normal cell homeostasis and function and for environmental adaption. Dynamic turnover of organelles is tightly controlled, with autophagy playing an essential role. Autophagy is a programmed process for efficient clearing of unwanted or damaged macromolecules or organelles, transporting them to vacuoles for degradation and recycling and thereby enhancing plant environmental plasticity. The specific autophagic engulfment of organelles requires activation of a selective autophagy pathway, recognition of the organelle by a receptor, and selective incorporation of the organelle into autophagosomes. While some of the autophagy machinery and mechanisms for autophagic removal of organelles is conserved across eukaryotes, plants have also developed unique mechanisms and machinery for these pathways. In this review, we discuss recent progress in understanding autophagy regulation in plants, with a focus on autophagic degradation of membrane-bound organelles. We also raise some important outstanding questions to be addressed in the future. 
    more » « less
  4. Despite theoretical benefits of replayability in educational games, empirical studies have found mixed evidence about the effects of replaying a previously passed game (i.e., elective replay) on students’ learning. Particularly, we know little about behavioral features of students’ elective replay process after experiencing failures (i.e., interruptive elective replay) and the relationships between these features and learning outcomes. In this study, we analyzed 5th graders’ log data from an educational game, ST Math, when they studied fractions—one of the most important but challenging math topics. We systematically constructed interruptive elective replay features by following students’ sequential behaviors after failing a game and investigated the relationships between these features and students’ post-test performance, after taking into account pretest performance and in-game performance. Descriptive statistics of the features we constructed revealed individual differences in the elective replay process after failures in terms of when to start replaying, what to replay, and how to replay. Moreover, a Bayesian multi-model linear regression showed that interruptive elective replay after failures might be beneficial for students if they chose to replay previously passed games when failing at a higher, more difficult level in the current game and if they passed the replayed games. 
    more » « less
  5. We propose an interacting contour stochastic gradient Langevin dynamics (IC-SGLD) sampler, an embarrassingly parallel multiple-chain contour stochastic gradient Langevin dynamics (CSGLD) sampler with efficient interactions. We show that ICSGLD can be theoretically more efficient than a single-chain CSGLD with an equivalent computational budget. We also present a novel random-field function, which facilitates the estimation of self-adapting parameters in big data and obtains free mode explorations. Empirically, we compare the proposed algorithm with popular benchmark methods for posterior sampling. The numerical results show a great potential of ICSGLD for large-scale uncertainty estimation tasks. 
    more » « less
  6. Recently, image-to-image translation (I2I) has met with great success in computer vision, but few works have paid attention to the geometric changes that occur during translation. The geometric changes are necessary to reduce the geometric gap between domains at the cost of breaking correspondence between translated images and original ground truth. We propose a novel geometry-aware semi-supervised method to preserve this correspondence while still allowing geometric changes. The proposed method takes a synthetic image-mask pair as input and produces a corresponding real pair. We also utilize an objective function to ensure consistent geometric movement of the image and mask through the translation. Extensive experiments illustrate that our method yields a 11.23% higher mean Intersection-Over-Union than the current methods on the downstream eye segmentation task. The generated image has a 15.9% decrease in Frechet Inception Distance indicating higher image quality. 
    more » « less