# Probabilistic Pedestrian Forecasts from a Handheld Phone: World-Frame Heat Maps, Visual-Inertial Height Drift, and Evaluation without Ground Truth

arXiv: https://arxiv.org/abs/2610.04736 · HTML: https://mlsanity.com/paper/2610.04736/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 31 of its 55 references that are in the map.
- First posted on arXiv 3 Oct 2026.

## Builds on (most cited first, 30 of 31)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [Rethinking Atrous Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1706.05587/) (2017)
- [ByteTrack: Multi-Object Tracking by Associating Every Detection Box](https://mlsanity.com/paper/2110.06864/) (2021)
- [Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks](https://mlsanity.com/paper/1803.10892/) (2018)
- [Accurate Uncertainties for Deep Learning Using Calibrated Regression](https://mlsanity.com/paper/1807.00263/) (2018)
- [AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting](https://mlsanity.com/paper/2103.14023/) (2021)
- [BoT-SORT: Robust Associations Multi-Pedestrian Tracking](https://mlsanity.com/paper/2206.14651/) (2022)
- [SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and Physical Constraints](https://mlsanity.com/paper/1806.01482/) (2018)
- [It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Prediction](https://mlsanity.com/paper/2004.02025/) (2020)
- [Stochastic Trajectory Prediction via Motion Indeterminacy Diffusion](https://mlsanity.com/paper/2203.13777/) (2022)
- [From Goals, Waypoints & Paths To Long Term Human Trajectory Forecasting](https://mlsanity.com/paper/2012.01526/) (2020)
- [Leapfrog Diffusion Model for Stochastic Trajectory Prediction](https://mlsanity.com/paper/2303.10895/) (2023)
- [Human Trajectory Forecasting in Crowds: A Deep Learning Perspective](https://mlsanity.com/paper/2007.03639/) (2020)
- [Safe Planning in Dynamic Environments using Conformal Prediction](https://mlsanity.com/paper/2210.10254/) (2022)
- [Peeking into the Future: Predicting Future Person Activities and Locations in Videos](https://mlsanity.com/paper/1902.03748/) (2019)
- [What the Constant Velocity Model Can Teach Us About Pedestrian Motion Prediction](https://mlsanity.com/paper/1903.07933/) (2019)
- [Future Person Localization in First-Person Videos](https://mlsanity.com/paper/1711.11217/) (2017)
- [Adaptive Conformal Prediction for Motion Planning among Dynamic Agents](https://mlsanity.com/paper/2212.00278/) (2022)
- [UCMCTrack: Multi-Object Tracking with Uniform Camera Motion Compensation](https://mlsanity.com/paper/2312.08952/) (2023)
- [Analyzing the Variety Loss in the Context of Probabilistic Trajectory Prediction](https://mlsanity.com/paper/1907.10178/) (2019)
- [Goal-driven Self-Attentive Recurrent Networks for Trajectory Prediction](https://mlsanity.com/paper/2204.11561/) (2022)
- [ADVIO: An authentic dataset for visual-inertial odometry](https://mlsanity.com/paper/1807.09828/) (2018)
- [OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets](https://mlsanity.com/paper/2010.00890/) (2020)
- [TBD Pedestrian Data Collection: Towards Rich, Portable, and Large-Scale Natural Pedestrian Data](https://mlsanity.com/paper/2309.17187/) (2023)
- [EgoTraj-Bench: Towards Robust Trajectory Prediction Under Ego-view Noisy Observations](https://mlsanity.com/paper/2510.00405/) (2025)
- [Indoor Future Person Localization from an Egocentric Wearable Camera](https://mlsanity.com/paper/2103.04019/) (2021)
- [T2FPV: Dataset and Method for Correcting First-Person View Errors in Pedestrian Trajectory Prediction](https://mlsanity.com/paper/2209.11294/) (2022)
- [Unified Human Localization and Trajectory Prediction with Monocular Vision](https://mlsanity.com/paper/2503.03535/) (2025)
- [HEADS-UP: Head-Mounted Egocentric Dataset for Trajectory Prediction in Blind Assistance Systems](https://mlsanity.com/paper/2409.20324/) (2024)

## Abstract (arXiv)

A pedestrian with a phone could be shown where nearby people will be in the next few seconds, if the forecast stays on the ground while the phone moves, is calibrated, and needs only a monocular camera and visual-inertial odometry (VIO). We build and evaluate such a system. People are detected, lifted onto the floor by ray-plane intersection, tracked in a gravity-aligned metric frame, and forecast as per-step probability maps by a small U-Net trained with a negative log-likelihood (NLL) loss on bird's-eye (SDD) and first-person (EgoTraj-Bench) trajectories. On handheld ADVIO recordings, vertical VIO drift and the user's own changes of level silently rescale monocular ground positions (by 87% within 90 s on one clip; on another, all tracks are lost for the last 31% of the clip); keeping the camera's height above the floor constant under a low-pass-filtered altitude avoids this, though it lags on escalators. On the SDD and EgoTraj-Bench test splits, the final forecaster lowers the NLL at 4.8 s by 1.51 and 1.37 nats relative to a fitted constant-velocity Gaussian. Lacking ground truth for people in handheld video, we score forecasts against the tracker's own later raw measurements. In an internally pre-registered evaluation on seven held-out clips, the final forecaster's NLL is lower than the benchmark-fitted baseline's at 1.2, 2.4 and 4.8 s (by 0.17, 0.23 and 0.47 nats; 95% intervals over people exclude zero), but by less than half as much as on the development clips. Exploratory analyses cut both ways: resampling clips instead of people widens the intervals to include zero at 1.2 and 2.4 s, and once both forecasters are recalibrated on the development clips the network is significantly better only at 1.2 s; but two clips run with ADVIO's reference poses favour the network much more when re-run with the phone's own poses. We discuss what such self-consistency scores can and cannot show.