Follow Everything: A Leader-Following and Obstacle Avoidance Framework with Goal-Aware Adaptation

Boyi Liu; Dimitrios Kanoulas; Jianhao Jiao; Qianyi Zhang; Shijian Ma

REVIEW 4 major objections 5 minor 26 references

Follow Everything: A Leader-Following and Obstacle Avoidance Framework with Goal-Aware Adaptation

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Legged robot follows anything, even lost leaders, with 96.9% success

desk verdict A solid, incremental leader-following paper whose distance-frame buffer is the real novelty; the ablation evidence is decent, but the confidence-score assumption and the abstract's mixed-baseline percentages need a referee's attention. read the letter →

arxiv 2504.19399 v5 pith:7ZZ6BX7A submitted 2025-04-28 cs.RO

Qianyi Zhang , Shijian Ma , Boyi Liu , Jianhao Jiao , Dimitrios Kanoulas This is my paper · ORCID

classification cs.RO

keywords leaderfollowingleggedrobotssegmentationmodelre-identificationdistanceframebuffergraph-basedplanninggoal-awareadaptationobstacleavoidance

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a single leader-following framework that replaces category detectors with a segmentation model, so the leader can be any object — a person, a wheeled robot, a quadrotor, or a stop sign. Its central claim is that a distance frame buffer, which stores the highest-confidence leader embedding for each distance range, prevents partial-view features captured just before the leader leaves the field of view from corrupting re-identification. Combined with a goal-aware state machine (chasing, following, planning, retreating, switching) and a graph-based time-optimal planner, the framework lets a legged robot follow arbitrary leaders and recover after temporarily losing sight of them. In simulation across four scenarios and 160 tests, it reports a 96.9% follow success rate, a 10.7% average leader-loss time ratio, a 1.8% collision rate, and a 2.0 m average leader–follower distance, outperforming both baselines and ablations on every metric.

What carries the argument

The central object is the distance frame buffer: an array of $n_2$ slots, where slot $i$ stores the leader embedding with the highest confidence score $S(\eta_t)$ among all historical embeddings whose follower–leader distance falls in the interval $[(i-1)\Delta d, i\Delta d]$. It works alongside a temporal memory buffer to bias segmentation toward a trustworthy reference at the current distance. The second piece of machinery is the goal-aware adaptation that, for each state, defines goal sets (lines, circles, or a fixed point) and velocity constraints fed into a graph-based trajectory optimizer: obstacle clusters become nodes, shortest collision-free connections become edges, and the planner enumerates detour trajectories, prunes them by homotopy class, and solves a time-optimal optimization to select the fastest feasible path.

What would settle it

In the playground scenario, record leader embeddings during partial occlusion near the field-of-view edge, then replace the distance-frame-buffer selection rule with "store the most recent embedding per distance bin" instead of "highest confidence"; if re-identification success after reappearance does not drop, then confidence-based selection is not load-bearing for the reported gains.

Watch

Extended reading notes

Core claim

The paper's discovery is that leader re-identification after visual loss is improved by partitioning the memory of leader appearances by distance. Instead of maintaining only a single temporal buffer of recent embeddings (which the paper argues can be dominated by low-quality, partial-view features from just before the leader exits the field of view), the distance frame buffer keeps one high-confidence embedding per equal distance interval between follower and leader. When the leader reappears, the per-distance memory supplies a more representative embedding for matching. The planning side contributes a state machine that sets different goal sets and speed constraints for each interaction mode, and a graph-based planner that generates homotopy-distinct candidate trajectories and optimizes the fastest one under those constraints. The net claim is that this combination yields substantially higher follow success and lower collision rates than the compared baselines in simulation.

Load-bearing premise

The central claim rests on the segmentation model's confidence score telling which stored embedding will best re-identify the leader later, even when the leader was only partly visible when that embedding was captured; if confidence is miscalibrated, the distance buffer may store the very features that cause re-identification failure.

Editorial extensions

If this is right

A robot following a person or another robot can keep re-identifying the target after occlusions or after the target leaves the camera view, as long as a high-confidence embedding was stored at a similar distance.
Switching to a speed-maximizing "planning" state toward the leader's last known pose reduces the chance of permanently losing the leader, compared to stopping or replicating the leader's path.
The goal-line and goal-circle constraint formulation lets the planner slide the trajectory endpoint along feasible goal sets, producing time-optimal paths without requiring a fixed waypoint.
The same planner and state machine can handle qualitatively different interactions (far, near, lost, retreating, switched) without separate planning algorithms per scenario.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The distance-binned memory idea is not specific to leader-following; it could apply to any long-term object re-identification task where appearance varies with distance, such as warehouse tracking or persistent person following with intermittent views.
The confidence-score selection is untested under miscalibration: if the segmentation model's scores are not calibrated, the buffer might store worse embeddings than a recency-based or coverage-based scheme; an ablation storing the newest embedding per bin would isolate this.
Real-world results are demonstrated only qualitatively (a few indoor and outdoor clips); the simulation metrics are the only quantitative claims, so transferring them to field conditions with varied lighting, crowds, or uneven terrain remains an open testable question.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Two things to know before reading. First, this is an honest, incremental systems paper: the new bits are a distance-frame buffer for re-identifying a leader after losing it, and a goal-aware state machine that feeds constraints into a graph-based planner. Second, the simulation ablation supports the central claim, but the paper never validates the confidence score that the buffer relies on, and the abstract's improvement percentages are computed against different baselines per metric.

What's genuinely new: Eq. (2) keeps, per distance bin, the historical embedding with highest confidence. That's a simple but sensible idea for leader-following, where the leader's appearance changes with distance and the last frames before FOV loss are partial. The goal-aware adaptation with chasing/following/planning/retreating states is a reasonable way to handle the interaction. The graph planner is mostly inherited from the authors' own prior TEB work, but the integration is clean.

The paper does real work: four simulation scenarios, 10 replayed leader trajectories each, four trials per trajectory, 160 tests total, with two baselines and two ablations. The ablations show each component helps: removing the distance buffer drops success from 96.9% to 62.5%, removing the graph planner drops it to 81.3%. That is meaningful evidence.

Soft spots, in proportion. The buffer's design depends on S(eta_t) being a trustworthy quality score. The paper's own introduction says partial low-quality features can dominate bootstrapping; the per-bin argmax only fixes that if, within each bin, the highest-S embedding is actually clean. No calibration or validation of S is provided. The stress-test also notes the bin index uses the centroid of visible leader points, which is biased under partial visibility. That's a real concern, but not fatal: the distance bins are coarse, and the overall buffer still improves performance empirically. I'd want a referee to ask for an ablation that isolates per-bin selection from simply keeping more embeddings.

Other soft spots: no baseline tuning details, no code or video URL in the text (the abstract promises a website), and the real-world results are qualitative. The abstract's numbers are computed against different baselines per metric - 75.1% success vs Alaa, 0.4 m distance vs SA-MPC - which reads as cherry-picking even if unintentional.

Who this is for: robotics researchers working on human/robot following, especially on legged platforms. It deserves serious review: the idea is plausible, the experiments are reproducible in structure, and the flaws are fixable. I'd send it to peer review with a request for code, a calibration check on the confidence score, and an ablation that teases apart the buffer's components.

Referee Report

4 major / 5 minor

Summary. The paper proposes a unified leader-following framework that combines a SAM2-based segmentation module with two memory mechanisms (a temporal memory buffer and a newly introduced distance frame buffer), a goal-aware state machine (chasing, following, planning, retreating, switching), and a graph-based trajectory planner. The method is evaluated in Gazebo using replayed leader trajectories across four scenarios with two external baselines (Alaa and SA-MPC) and two ablations, plus qualitative real-world demonstrations with a legged robot. The headline simulation results are a 96.9% follow success rate, a 10.7% average leader loss time ratio, a 1.8% collision rate, and a 2.0 m average distance, reported as best among all compared methods.

Significance. If the empirical results hold, the paper provides a useful integration for open-set leader following with reacquisition after the leader leaves the field of view. The simulation protocol, which replays recorded leader trajectories multiple times, is reasonable, and the ablation table indicates that both the distance frame buffer and the graph-based planner contribute to the reported improvement. The breadth of leader types (human, ground robot, UAV, legged robot, stop sign) is a strength, and the promise of released code and video is valuable. The work is primarily an integration of existing perception and planning components with a new distance-conditional memory design; its novelty is moderate. The main weaknesses are in evaluation rigor: inconsistent abstract comparisons, missing variance and baseline-tuning details, and an unvalidated confidence-selection mechanism in the distance frame buffer.

major comments (4)

[Abstract and Table I] The four improvement percentages in the abstract are not computed against a single baseline. Specifically, 75.1% is the follow-success gain over Alaa (96.9 - 21.8), 13.1% is the loss-time reduction over Alaa (23.8 - 10.7), 65.1% is the collision-rate reduction over Alaa (66.9 - 1.8), and 0.4 m is the distance reduction over SA-MPC (2.4 - 2.0). Presenting these together as 'improvements' obscures the fact that the reference method changes from metric to metric. Please either report each percentage with its explicit baseline or use one consistent baseline throughout. Additionally, the table would be much stronger with per-scenario results and variance information.
[Section II-B, Eq. (2)] The distance frame buffer stores, for each distance bin, the historical embedding with the highest confidence score S(eta_t), but the paper neither defines S nor validates that a high confidence score corresponds to a clean, full-body view of the leader. This is load-bearing because the Introduction argues that partial FOV-edge features are low-quality yet can dominate bootstrapping; if such a corrupted embedding receives a high confidence score and is stored in a distance bin, the buffer may reproduce the same failure it is designed to avoid. The ablation FE-N-DFB shows that adding the buffer helps in aggregate, but it does not isolate the confidence-selection rule from other differences such as buffer capacity or retention policy. Please add an experiment that simulates partial visibility near the FOV edge and checks whether the selected embeddings are in fact clean full-body views, or compare confidence-based selection against random selection within each bin.
[Section IV] The real-world experiments are entirely qualitative: each scenario is illustrated by a sequence of images, with no measured follow success rate, leader loss time, collision events, average distance, or number of trials. Since the abstract claims that simulations and real-world experiments demonstrate competitive improvements, the absence of quantitative real-world results leaves a large part of the claim unsupported. Please add a quantitative real-world evaluation (even a small table with per-scenario metrics) or explicitly relabel these as demonstrations rather than experiments supporting the performance claims.
[Section III, Table I] The simulation results are reported as point estimates without confidence intervals or standard deviations, even though the protocol replays 10 leader trajectories four times per scenario. This makes it difficult to judge whether the differences between Follow Everything and the baselines are statistically meaningful or dominated by trajectory-level variation. The paper also gives no details on how the baseline hyperparameters were tuned (e.g., PID gains for Alaa, MPC weights and horizons for SA-MPC); without a documented tuning protocol, a reader cannot rule out that the baselines are under-tuned. Please provide per-scenario statistics and a short description of how each baseline was configured.

minor comments (5)

[Section II-B, Eqs. (1)-(2)] The notation for the memory-buffer update is mathematically unclear: expressions such as 'arg max over forall eta_t cap eta_t notin [eta_1,...,eta_{i-1}]' are not standard set notation and do not define how the buffer is updated over time. Please rewrite these equations with explicit time indices and set definitions, and define where S(eta_t) comes from.
[Figure 7] The bar chart shows only means, while the text claims the method has the 'smallest variance' in average distance. Please add error bars or a violin/box plot so that the variance claim is visible and checkable.
[Title and Abstract] The title given in the first abstract ('Follow Everything: A Leader-Following and Obstacle Avoidance Framework with Goal-Aware Adaptation') differs from the title in the full text ('Follow Everything: Goal-Aware Adaptation and Graph-Based Planning Towards Arbitrary Leader Following'). The manuscript should use a single consistent title.
[Section II-D, Switching state] The text says a new leader is 'most commonly triggered by a large language model [21]', but reference [21] is a vision-language SAM variant, not a large language model. Please correct the citation or the description.
[Section II-C, Eq. (9)] The quantity f(tau*_{t-1}, tau) is called a 'similarity' but, as written, it can take values outside [0,1] depending on alpha and the number of obstacles. Please state the intended range or add a normalization that makes the similarity interpretation precise.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the empirical results and ablation comparisons stand on their own, and the cited prior planner work is external, not a load-bearing logical input.

full rationale

The paper's central claims are empirical: follow success rate, leader-loss time ratio, collision rate, and average distance are measured in simulation and real-world trials against external baselines (Alaa, SA-MPC) and ablations (FE-N-DFB, FE-N-GP). The distance frame buffer of Eq. (2) is a system component whose benefit is justified by the Table I ablation, not by defining the metric in terms of the component. The goal-aware adaptation and graph-based planner are constructive control and planning formulations (Eqs. (4)-(16)) with hand-tuned scaling factors; tuning choices are not circular reasoning. The graph construction cites prior works [22] and [23] by overlapping authors, but those are separately published, externally validated methods, and the present paper also states the optimization equations it uses; the citation is inheritance of a building block, not a self-referential proof of the current result. The confidence-score selection in Eqs. (1)-(2) creates a bootstrap loop that could be fragile, but that is a robustness and calibration risk, not a definitional equivalence between input and output. No step in the derivation reduces a predicted quantity to a fitted parameter or to the paper's own conclusion by construction.

Assumptions & free parameters 9 free parameters · 4 assumptions · 1 invented entities

The framework depends on a set of hand-tuned scaling factors and thresholds that are not reported in the manuscript, plus assumptions about the reliability of the segmentation confidence scores, the topological graph representation, and the leader position estimates. These are the unexamined inputs on which the empirical results rest.

free parameters (9)

α (Eq 10) = not reported
Scales the goal line length L_t in the chasing state; depends on costmap width and leader distance. No value or sensitivity analysis is given.
α (Eq 13) = not reported
Scales the NIS value to compute the safe distance D_t in the following state.
α (Eq 9) = not reported
Weights trajectory similarity vs time cost in the selection function; reused symbol with different meaning.
α1, α2 (Eq 15) = not reported
Scale leader velocity and distance terms in the following state's velocity limit Vmax_t.
Dmin, Dmax = not reported
Bounds for the dynamically adjusted safe distance in Eq (13).
Δd = not reported
Distance interval width for the distance frame buffer in Eq (2).
n1, n2 = not reported
Capacity of the temporal memory buffer and distance frame buffer, respectively.
Wmap = not reported
Costmap width used in Eq (10) to compute the goal line length.
ϵ = not reported
Small threshold for goal position and orientation constraints in Eqs (14) and (16).

assumptions (4)

domain assumption The segmentation model (cited as [21]) produces embeddings whose confidence scores S(η_t) are reliable indicators of match quality across different distances and partial views.
The distance frame buffer (Eq 2) selects the highest-confidence embedding per distance bin. If confidence is miscalibrated when the leader is partially visible at the FOV edge, the buffer stores corrupted embeddings and the re-identification claim fails. The paper does not validate this.
domain assumption The topological graph built from the costmap (following [22,23]) captures all relevant obstacle detour alternatives.
Section II-C assumes obstacle clusters as nodes and shortest connections as edges generate a sufficient set of candidate trajectories; this is inherited from prior work and not re-validated.
domain assumption The leader's position estimate, the average of the segmented point set from depth, is accurate enough for planning.
Section II-A takes the mean of Pleader as the leader position; errors from depth noise or mask leakage directly affect the goal sets and safe distance.
ad hoc to paper The Kalman filter's NIS value is a meaningful signal for adjusting the desired safe distance in the following state.
Eq (13) uses Clip(α NIS, Dmin, Dmax) with no prior validation; this is a heuristic introduced for this paper.

invented entities (1)

Distance frame buffer (B^D_t)
purpose: Stores the highest-confidence leader embedding for each distance interval to prevent low-quality partial-view embeddings from dominating the memory and to improve re-identification after the leader reappears.
Introduced in Eq (2). Its efficacy is only evidenced by the paper's own ablation (FE vs FE-N-DFB); it makes no independently falsifiable prediction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Follow Everything: A Leader-Following and Obstacle Avoidance Framework with Goal-Aware Adaptation." pith.science (2026). https://pith.science/paper/7ZZ6BX7A

@misc{pith2026250419399,
  author       = {Pith},
  title        = {Pith review of: Follow Everything: A Leader-Following and Obstacle Avoidance Framework with Goal-Aware Adaptation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7ZZ6BX7A}},
  note         = {Machine review of arXiv:2504.19399}
}

read the original abstract

Robust and flexible leader-following is a critical capability for robots to integrate into human society. While existing methods struggle to generalize to leaders of arbitrary form and often fail when the leader temporarily leaves the robot's field of view, this work introduces a unified framework addressing both challenges. First, traditional detection models are replaced with a segmentation model, allowing the leader to be anything. To enhance recognition robustness, a distance frame buffer is implemented that stores leader embeddings at multiple distances, accounting for the unique characteristics of leader-following tasks. Second, a goal-aware adaptation mechanism is designed to govern robot planning states based on the leader's visibility and motion, complemented by a graph-based planner that generates candidate trajectories for each state, ensuring efficient following with obstacle avoidance. Simulations and real-world experiments with a legged robot follower and various leaders (human, ground robot, UAV, legged robot, stop sign) in both indoor and outdoor environments show competitive improvements in follow success rate, reduced visual loss duration, lower collision rate, and decreased leader-follower distance.

Figures

Figures reproduced from arXiv: 2504.19399 by the authors.

**Figure 2.** Illustration of the proposed follow everything framework. Given a leader prompt and an RGB image, the leader mask is segmented with the aid of both a memory buffer and a distance frame buffer. The segmented leader is then filtered from raw laser point clouds to produce a leader point set, which is passed to the goal-aware adaptation to dynamically adjust parameters and provide goal sets and constraints according to … view at source ↗

**Figure 3.** Illustration of goal-aware adaptation and graph-based planner in a demo. (a) The leader is segmented after selecting points. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: Simulation for following a mobile robot in a playground. (a) The leader can be robustly segmented even when its color closely resembles the [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Simulation for following a pedestrian in a factory. (a–b) The robot follows the leader when partially observed. (c) Upon receiving a prompt [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

**Figure 6.** Figure 6: Simulation for following a UAV in a forest. (a) After the UAV [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

**Figure 7.** Figure 7: Average robot-leader distance in 160 tests: Alaa 3.3 m, SA-MPC [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 9.** Figure 9: Illustration of real-world experiments. (a-c) While following a pedestrian, an appropriate safe distance allows the robot to rotate in place and [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 21 canonical work pages

[1]

P-agnav: Range view-based autonomous navigation system for cornfields,

K. Kim, A. Deb, and D. J. Cappelleri, “P-agnav: Range view-based autonomous navigation system for cornfields,” IEEE Robotics and Automation Letters, vol. 10, no. 4, pp. 3366–3373, 2025

work page 2025
[2]

Gpt-driven gestures: Leveraging large language models to generate expressive robot motion for enhanced human-robot interaction,

L. Roy, E. A. Croft, A. Ramirez, and D. Kuli ´c, “Gpt-driven gestures: Leveraging large language models to generate expressive robot motion for enhanced human-robot interaction,” IEEE Robotics and Automa- tion Letters, vol. 10, no. 5, pp. 4172–4179, 2025

work page 2025
[3]

Legged robot- aided 3d tunnel mapping via residual compensation and anomaly detection,

X. Zhang, Z. Huang, Q. Li, R. Wang, and B. Zhou, “Legged robot- aided 3d tunnel mapping via residual compensation and anomaly detection,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 214, pp. 33–47, 2024

work page 2024
[4]

Unlocking aerobatic potential of quadcopters: Autonomous freestyle flight generation and execution,

M. Wang, Q. Wang, Z. Wang, Y . Gao, J. Wang, C. Cui, Y . Li, Z. Ding, K. Wang, C. Xu, and F. Gao, “Unlocking aerobatic potential of quadcopters: Autonomous freestyle flight generation and execution,” Science Robotics, vol. 10, no. 101, p. eadp9905, 2025

work page 2025
[5]

Leader-follower formation tracking control of mobile robots: A visual observer-based approach,

S. Huang, H. Zeng, W. Lan, and X. Yu, “Leader-follower formation tracking control of mobile robots: A visual observer-based approach,” IEEE Transactions on Control Systems Technology , pp. 1–9, 2025

work page 2025
[6]

Learning robust autonomous navigation and locomotion for wheeled- legged robots,

J. Lee, M. Bjelonic, A. Reske, L. Wellhausen, T. Miki, and M. Hutter, “Learning robust autonomous navigation and locomotion for wheeled- legged robots,” Science Robotics, vol. 9, no. 89, p. eadi9641, 2024

2024
[7]

A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas,

J. Terven, D.-M. Córdova-Esparza, and J.-A. Romero-González, “A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas,” Machine learning and knowledge extraction, vol. 5, no. 4, pp. 1680–1716, 2023

work page 2023
[8]

Center-based 3d object detec- tion and tracking,

T. Yin, X. Zhou, and P. Krahenbuhl, “Center-based 3d object detec- tion and tracking,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 11 784–11 793

2021

Show all 26 references

[9]

Hybrid object detection and tracking for cooperative perception using 3d lidar,

Z. Meng, X. Xia, R. Xu, W. Liu, and J. Ma, “Hybrid object detection and tracking for cooperative perception using 3d lidar,” IEEE Trans- actions on Intelligent Vehicles , vol. 8, no. 8, pp. 4069–4080, 2023

2023
[10]

Follow anything: Open-set detection, tracking, and following in real-time,

A. Maalouf, N. Jadhav, K. M. Jatavallabhula, M. Chahine, D. M. V ogt, R. J. Wood, A. Torralba, and D. Rus, “Follow anything: Open-set detection, tracking, and following in real-time,” IEEE Robotics and Automation Letters, vol. 9, no. 4, pp. 3283–3290, 2024

2024
[11]

Safe and robust human follow- ing for mobile robots based on self-avoidance mpc in crowded corridor scenarios,

Y . Song, Q. Zhang, Z. Hu, and J. Liu, “Safe and robust human follow- ing for mobile robots based on self-avoidance mpc in crowded corridor scenarios,” in 2023 IEEE International Conference on Robotics and Biomimetics (ROBIO), 2023, pp. 1–6

2023
[12]

The human gaze helps robots run bravely and efficiently in crowds,

Q. Zhang, Z. Hu, Y . Song, J. Pei, and J. Liu, “The human gaze helps robots run bravely and efficiently in crowds,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 7540–7546

2023
[13]

Efficient motion planning based on kinodynamic model for quadruped robots following persons in confined spaces,

Z. Zhang, J. Yan, X. Kong, G. Zhai, and Y . Liu, “Efficient motion planning based on kinodynamic model for quadruped robots following persons in confined spaces,” IEEE/ASME Transactions on Mechatron- ics, vol. 26, no. 4, pp. 1997–2006, 2021

1997
[14]

Diffusion-based 2d path planner applied on legged robots,

J. Liu, M. Stamatopoulou, and D. Kanoulas, “Diffusion-based 2d path planner applied on legged robots,” in IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 9264–9270

2024
[15]

Direct point robot navigation with end- to-end model-based learning,

R. Han, S. Wang, S. Wang, Z. Zhang, J. Chen, S. Lin, C. Li, C. Xu, Y . C. Eldar, Q. Hao et al. , “Direct point robot navigation with end- to-end model-based learning,” IEEE Transactions on Robotics , 2025

2025
[16]

A quadruped robot obstacle avoidance and personnel following strategy based on ultra-wideband and three-dimensional laser radar,

Z. Li, B. Li, Q. Liang, W. Liu, L. Hou, and X. Rong, “A quadruped robot obstacle avoidance and personnel following strategy based on ultra-wideband and three-dimensional laser radar,” Inter- national Journal of Advanced Robotic Systems , vol. 19, no. 4, p. 17298806221114705, 2022

2022
[17]

Efficient volumetric mapping of multi-scale environments using wavelet-based compres- sion,

V . Reijgwart, C. Cadena, R. Siegwart, and L. Ott, “Efficient volumetric mapping of multi-scale environments using wavelet-based compres- sion,” arXiv preprint arXiv:2306.01279 , 2023

2023 arXiv
[18]

Obstacle-avoidant leader following with a quadruped robot,

C. Scheidemann, L. Werner, V . Reijgwart, A. Cramariuc, J. Chomarat, J.-R. Chiu, and M. Hutter, “Obstacle-avoidant leader following with a quadruped robot,” arXiv preprint arXiv:2410.00572 , 2024

2024 arXiv
[19]

Human following robot using ultrasonic sensor,

A. Tripathi, M. A. Khan, A. Pandey, P. Yadav, and A. K. Sharma, “Human following robot using ultrasonic sensor,” in 2021 3rd In- ternational Conference on Advances in Computing, Communication Control and Networking (ICAC3N) . IEEE, 2021, pp. 764–770

2021
[20]

Kalman filter and its application,

Q. Li, R. Li, K. Ji, and W. Dai, “Kalman filter and its application,” in 2015 8th International Conference on Intelligent Networks and Intelligent Systems (ICINIS) , 2015, pp. 74–77

2015
[21]

Early vision-language fusion for text-prompted segment anything model,

Y . Zhang, T. Cheng, R. Hu, L. Liu, H. Liu, L. Ran, X. Chen, W. Liu, and X. Wang, “Early vision-language fusion for text-prompted segment anything model,” arXiv preprint arXiv:2406.20076 , 2024

2024 arXiv
[22]

Stc-teb: Spatial- temporally complete trajectory generation based on incremental opti- mization,

Z. Zhu, Q. Zhang, Y . Song, Y . Yang, and J. Liu, “Stc-teb: Spatial- temporally complete trajectory generation based on incremental opti- mization,” IEEE Robotics and Automation Letters , 2024

2024
[23]

Ga-teb: Goal- adaptive framework for efficient navigation based on goal lines,

Q. Zhang, W. Luo, Z. Zhang, Y . Wang, and J. Liu, “Ga-teb: Goal- adaptive framework for efficient navigation based on goal lines,” 2024

2024
[24]

Topology-driven parallel trajectory optimization in dynamic environ- ments,

O. de Groot, L. Ferranti, D. M. Gavrila, and J. Alonso-Mora, “Topology-driven parallel trajectory optimization in dynamic environ- ments,” IEEE Transactions on Robotics , vol. 41, pp. 110–126, 2025

2025
[25]

Search-based path planning with homotopy class constraints,

S. Bhattacharya, “Search-based path planning with homotopy class constraints,” in Proceedings of the AAAI conference on artificial intelligence, vol. 24, no. 1, 2010, pp. 1230–1237

2010
[26]

Integrated online trajec- tory planning and optimization in distinctive topologies,

C. Rösmann, F. Hoffmann, and T. Bertram, “Integrated online trajec- tory planning and optimization in distinctive topologies,” Robotics and Autonomous Systems, vol. 88, pp. 142–153, 2017

2017

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

P-agnav: Range view-based autonomous navigation system for cornfields,

K. Kim, A. Deb, and D. J. Cappelleri, “P-agnav: Range view-based autonomous navigation system for cornfields,” IEEE Robotics and Automation Letters, vol. 10, no. 4, pp. 3366–3373, 2025

work page 2025

[2] [2]

Gpt-driven gestures: Leveraging large language models to generate expressive robot motion for enhanced human-robot interaction,

L. Roy, E. A. Croft, A. Ramirez, and D. Kuli ´c, “Gpt-driven gestures: Leveraging large language models to generate expressive robot motion for enhanced human-robot interaction,” IEEE Robotics and Automa- tion Letters, vol. 10, no. 5, pp. 4172–4179, 2025

work page 2025

[3] [3]

Legged robot- aided 3d tunnel mapping via residual compensation and anomaly detection,

X. Zhang, Z. Huang, Q. Li, R. Wang, and B. Zhou, “Legged robot- aided 3d tunnel mapping via residual compensation and anomaly detection,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 214, pp. 33–47, 2024

work page 2024

[4] [4]

Unlocking aerobatic potential of quadcopters: Autonomous freestyle flight generation and execution,

M. Wang, Q. Wang, Z. Wang, Y . Gao, J. Wang, C. Cui, Y . Li, Z. Ding, K. Wang, C. Xu, and F. Gao, “Unlocking aerobatic potential of quadcopters: Autonomous freestyle flight generation and execution,” Science Robotics, vol. 10, no. 101, p. eadp9905, 2025

work page 2025

[5] [5]

Leader-follower formation tracking control of mobile robots: A visual observer-based approach,

S. Huang, H. Zeng, W. Lan, and X. Yu, “Leader-follower formation tracking control of mobile robots: A visual observer-based approach,” IEEE Transactions on Control Systems Technology , pp. 1–9, 2025

work page 2025

[6] [6]

Learning robust autonomous navigation and locomotion for wheeled- legged robots,

J. Lee, M. Bjelonic, A. Reske, L. Wellhausen, T. Miki, and M. Hutter, “Learning robust autonomous navigation and locomotion for wheeled- legged robots,” Science Robotics, vol. 9, no. 89, p. eadi9641, 2024

2024

[7] [7]

A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas,

J. Terven, D.-M. Córdova-Esparza, and J.-A. Romero-González, “A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas,” Machine learning and knowledge extraction, vol. 5, no. 4, pp. 1680–1716, 2023

work page 2023

[8] [8]

Center-based 3d object detec- tion and tracking,

T. Yin, X. Zhou, and P. Krahenbuhl, “Center-based 3d object detec- tion and tracking,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 11 784–11 793

2021