Yasamin Mostofi

Anurag Pallaprolu, Winston Hurst, Yasamin Mostofi

In this paper, we present mmFlux: a novel framework for extracting underlying crowd motion patterns and inferring crowd semantics using mmWave radar. First, our proposed signal processing pipeline combines optical flow estimation concepts from vision with novel statistical and morphological noise filtering. This approach generates high-fidelity mmWave flow fields-compact 2D vector representations of crowd motion. We then introduce a novel approach that transforms these fields into directed geometric graphs. In these graphs, edges capture dominant flow currents, vertices mark crowd splitting or merging, and flow distribution is quantified across edges. Finally, we show that analyzing the local Jacobian and computing the corresponding curl and divergence enables extraction of key crowd semantics for both structured and diffused crowds. We conduct 21 experiments on crowds of up to 20 people across 3 areas, using commodity mmWave radar. Our framework achieves high-fidelity graph reconstruction of the underlying flow structure, even for complex crowd patterns, demonstrating strong spatial alignment and precise quantitative characterization of flow split ratios. Finally, our curl and divergence analysis accurately infers key crowd semantics, e.g., abrupt turns, boundaries where flow directions shift, dispersions, and gatherings. Overall, these findings validate mmFlux, underscoring its potential for various crowd analytics applications.

Chitra R. Karanam, Yasamin Mostofi

In this paper, we consider the problem of wireless channel prediction, where we are interested in predicting the channel quality at unvisited locations in an area of interest, based on a small number of prior received power measurements collected by an unmanned vehicle in the area. We propose a new framework for channel prediction that can not only predict the detailed variations of the received power, but can also predict the detailed makeup of the wireless rays (i.e., amplitude, angle-of-arrival, and phase of all the incoming paths). More specifically, we show how an enclosure-based robotic route design ensures that the received power measurements at the prior measurement locations can be utilized to fully predict detailed ray parameters at unvisited locations. We then show how to first estimate the detailed ray parameters at the prior measurement route and then fully extend them to predict the detailed ray makeup at unvisited locations in the workspace. We experimentally validate our proposed framework through extensive real-world experiments in three different areas, and show that our approach can accurately predict the received channel power and the detailed makeup of the rays at unvisited locations in an area, considerably outperforming the state-of-the-art in wireless channel prediction.

Arjun Muralidharan, Yasamin Mostofi

In this paper, we consider a scenario where a robot needs to establish connectivity with a remote operator or another robot, as it moves along a path. We are interested in answering the following question: what is the distance traveled by the robot along the path before it finds a connected spot? More specifically, we are interested in characterizing the statistics of the distance traveled along the path before it gets connected, in realistic channel environments experiencing path loss, shadowing and multipath effects. We develop an exact mathematical analysis of these statistics for straight-line paths and also mathematically characterize a more general space of paths (beyond straight paths) for which the analysis holds, based on the properties of the path such as its curvature. Finally, we confirm our theoretical analysis using extensive numerical results with real channel parameters from downtown San Francisco.

Ekta Prashnani, Hong Cai, Yasamin Mostofi et al.

The ability to estimate the perceptual error between images is an important problem in computer vision with many applications. Although it has been studied extensively, however, no method currently exists that can robustly predict visual differences like humans. Some previous approaches used hand-coded models, but they fail to model the complexity of the human visual system. Others used machine learning to train models on human-labeled datasets, but creating large, high-quality datasets is difficult because people are unable to assign consistent error labels to distorted images. In this paper, we present a new learning-based method that is the first to predict perceptual image error like human observers. Since it is much easier for people to compare two given images and identify the one more similar to a reference than to assign quality scores to each, we propose a new, large-scale dataset labeled with the probability that humans will prefer one image over another. We then train a deep-learning model using a novel, pairwise-learning framework to predict the preference of one distorted image over the other. Our key observation is that our trained network can then be used separately with only one distorted image and a reference to predict its perceptual error, without ever being trained on explicit human perceptual-error labels. The perceptual error estimated by our new metric, PieAPP, is well-correlated with human opinion. Furthermore, it significantly outperforms existing algorithms, beating the state-of-the-art by almost 3x on our test set in terms of binary error rate, while also generalizing to new kinds of distortions, unlike previous learning-based methods.

Arjun Muralidharan, Yasamin Mostofi

Consider a general path planning problem of a robot on a graph with edge costs, and where each node has a Boolean value of success or failure (with respect to some task) with a given probability. The objective is to plan a path for the robot on the graph that minimizes the expected cost until success. In this paper, it is our goal to bring a foundational understanding to this problem. We start by showing how this problem can be optimally solved by formulating it as an infinite horizon Markov Decision Process, but with an exponential space complexity. We then formally prove its NP-hardness. To address the space complexity, we then propose a path planner, using a game-theoretic framework, that asymptotically gets arbitrarily close to the optimal solution. Moreover, we also propose two fast and non-myopic path planners. To show the performance of our framework, we do extensive simulations for two scenarios: a rover on Mars searching for an object for scientific studies, and a robot looking for a connected spot to a remote station (with real data from downtown San Francisco). Our numerical results show a considerable performance improvement over existing state-of-the-art approaches.