Pith. sign in

REVIEW 3 major objections 6 minor 36 references

EsaacSim: A Multimodal Event Camera Add-on for NVIDIA Isaac Sim

T0 review · 3 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper claims that an add-on for the Isaac Sim robotics simulator can generate synchronized event, RGB, APS, depth, and IMU streams online, with effective event rates up to 960 Hz, using motion-guided interpolation and a frame-based…

desk verdict Useful Isaac Sim event-camera integration, but the motion-guided interpolation equation is time-reversed and the effective-rate claims are not trustworthy until that is fixed and real-camera validation appears. read the letter →

arxiv 2608.08522 v2 pith:C677N3LO submitted 2026-08-09 cs.RO

Ignacio Bugueno-Cordova , Malte Kuhlmann , Nicolás Navarro-Guerrero , Miguel Campusano , Rodrigo Verschae This is my paper · ORCID
classification cs.RO
keywords EventcamerasimulationIsaacSimNeuromorphicvisionRoboticsROS2BayerRGGBeventsMotion-guidedinterpolationSyntheticdatageneration
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

This paper claims that an add-on for the Isaac Sim robotics simulator can produce synchronized event, RGB, APS, depth, and IMU streams online, directly from the simulator's existing render pipeline. The central move is to attach a frame-based log-intensity event model to rendered HDR images, with motion-guided synthesis of intermediate frames that raises the effective event rate to up to 960 Hz without increasing the render frequency. Benchmarks across five event-camera resolutions report 6.98–27.28 ms per rendered-frame interval for grayscale events and 7.58–29.16 ms for Bayer RGGB events, using less than 400 MB of extra GPU memory on a consumer-grade graphics card. If these numbers hold, event-based perception such as SLAM, optical flow, and object detection can be developed and tested in a photorealistic simulator with closed-loop ROS 2 integration, rather than on scarce physical event cameras.

What carries the argument

The load-bearing mechanism is the pair formed by a frame-based log-intensity event model and motion-guided frame-gap synthesis. The event model computes $L_k(x,y)=\log(S_k(x,y)+1)$, keeps a per-pixel reference level $L_{\mathrm{ref}}$, and emits ON/OFF events when $\Delta L_k(x,y)$ reaches $C_{\mathrm{ON}}$ or $C_{\mathrm{OFF}}$. The synthesis step takes two consecutive rendered frames with the renderer's dense motion-vector field $v(x,y)=(u(x,y),v(x,y))$ and forms intermediate samples by backward warping $I_{k,n}(x,y)=I_k(x-\alpha u(x,y), y-\alpha v(x,y))$ with bilinear interpolation, where $\alpha=n/N$; timestamps are placed uniformly between the two frame times. This lets the simulator produce high-temporal-resolution event streams from a low render cadence, reusing the same event model on real and synthesized frames.

What would settle it

Record a real scene with an event camera and replay the same camera trajectory and lighting in EsaacSim; if the generated events miss systematic structure such as high-dynamic-range details, motion blur, or noise, or if downstream perception performance diverges markedly from real data, the central usefulness claim would be undermined. A concrete quantitative check is to compare event-rate distributions and contrast-threshold histograms between the real and simulated streams under identical motion.

Watch

Extended reading notes

Core claim

EsaacSim's central claim is that online, configurable event camera simulation can be layered on top of a GPU-accelerated robotics simulator without altering the renderer. It consumes the camera's HDR render product, converts it to log-intensity (grayscale, weighted luminance, or Bayer RGGB samples), and emits an asynchronous event when the accumulated log-intensity change at a pixel crosses a positive or negative contrast threshold. Sensor parameters such as thresholds, threshold mismatch, refractory period, latency, timestamp quantization, shot noise, and leak noise are runtime-configurable per camera. To increase temporal resolution, the add-on synthesizes intermediate images by backward-warping the current rendered image along the renderer's motion vectors for $N$ fractional steps between frames; both rendered and synthesized images go through the same event model. The result is a unified asynchronous event stream with effective sampling rates up to 960 Hz, delivered alongside RGB, APS, depth, and IMU data through native ROS 2 interfaces.

Load-bearing premise

The load-bearing premise is that thresholded log-intensity changes computed from rendered HDR frames, including the motion-warped synthesized frames, faithfully reproduce the event stream a real event camera would produce for the same scene; the paper reports no quantitative comparison against real event camera recordings.

Editorial extensions

If this is right

  • A 30 Hz render loop can yield event streams at effective sampling rates of 240, 480, or 960 Hz, so high-speed motion can be studied without raising the rendering cost.
  • Synchronized RGB, APS, depth, IMU, and event data are available through native ROS 2 messages, making the add-on directly usable by robot perception and control stacks.
  • Five common event-camera resolutions (DVS128, DAVIS240, ATIS, DAVIS346, GenX320) are supported with configurable thresholds, latency, refractory periods, and noise, so multiple sensor profiles can be emulated in one scene.
  • The measured per-frame event-generation time stays below about 30 ms on a consumer-grade graphics card with under 400 MB of extra GPU memory, which is consistent with online operation alongside the simulator's own rendering.
  • Both grayscale and Bayer RGGB event generation are handled by the same backend, so color event camera configurations can be simulated without a separate pipeline.

Reading between the lines

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

  • A natural next test would be to record the same scene with a real event camera and compare event-rate statistics, threshold distributions, and downstream odometry error; the paper's public release would make this validation possible.
  • If the motion-warped interpolation proves faithful, the same synthesis approach could be reused for other high-frame-rate camera emulation and for data augmentation in event-based learning.
  • The architecture's separation between rendering and event backend suggests the physics-based backend currently stubbed out could later be swapped in without changing the ROS 2 interfaces, but no evidence yet shows it would match real sensor behavior.
  • Closed-loop training of event-based policies in simulation could become common practice before hardware deployment, assuming fidelity holds; this would reduce reliance on expensive physical event cameras.
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.

Referee Report

3 major / 6 minor

Summary. The paper presents EsaacSim, an add-on for NVIDIA Isaac Sim that generates online event-camera streams from rendered images, supporting grayscale and Bayer RGGB event modes together with synchronized RGB, APS, depth, and IMU outputs exposed through ROS 2. The event generation uses a standard frame-based log-intensity threshold model, and a motion-guided frame-gap synthesis method is proposed to increase the effective temporal sampling rate up to 960 Hz without increasing the rendering frequency. Experiments on an RTX 4060 report event-generation times of 6.98--27.28 ms per rendered-frame interval for grayscale and 7.58--29.16 ms for Bayer RGGB across five sensor resolutions, with less than 400 MB of additional GPU memory. The authors acknowledge that the simulated event streams are not yet quantitatively validated against real event camera data and that the physics-based backend is not implemented.

Significance. If the method is correct and the generated event streams are faithful, EsaacSim would fill a practical gap: an integrated, ROS 2-ready multimodal event camera simulator inside Isaac Sim. The paper makes a concrete engineering contribution by combining configurable sensor models, Bayer RGGB event generation, synchronized multimodal outputs, and a consumer-GPU performance evaluation. The performance measurements, while limited to a single GPU and without repeated-run statistics, are useful and clearly reported. The central algorithmic question, however, is the correctness of the motion-guided frame-gap synthesis, and the lack of any fidelity validation against real event cameras leaves the usefulness of the generated data for robotics perception unestablished.

major comments (3)
  1. [§4.2, Eqs. (4) and (5)] The motion-guided frame-gap synthesis appears internally time-reversed under the standard convention for backward motion vectors. If v(x,y) points from a pixel's current position to its previous position, then Eq. (4) gives I_k(x) at α=0 and I_k(x−v), i.e., the previous frame's content, at α=1. Eq. (5), however, assigns α=0 to t_{k−1} and α=1 to t_k. The synthesized intermediate images are therefore ordered in reverse along the interpolation interval. For N=2 the single midpoint is symmetric and the error is hidden, but for the reported N=8, 16, and 32, early synthesized samples contain content near the current frame while late samples contain content near the previous frame. Because each synthesized frame is passed through the same log-intensity event model, the events inherit incorrect timestamps and reversed polarity ordering for a given direction of scene motion. This directly affects the paper's central claim of effective temporal sampling rates up to 960 Hz. Please correct Eq. (4) (e.g., use I_k(x+(1−α)v) under the current-to-previous convention, or equivalently I_{k−1}(x+αv)) or, if the implementation already uses the corrected form, amend the equation to match the implementation and state the motion-vector convention explicitly.
  2. [§5.3 and Conclusions] The paper reports no quantitative comparison of the generated event streams against real event camera recordings. The authors themselves state in Section 5.3 that 'establishing this correspondence is necessary to assess the fidelity of the generated event streams.' Since the stated purpose of the tool is to support robotics research with synthetic event data, computational efficiency alone does not establish that the events are useful. I request at least a basic fidelity evaluation: for example, known-motion scenes with quantitative event-rate and polarity checks, comparison of event counts versus contrast threshold against a real DAVIS or GenX sensor, or comparison with established simulators such as ESIM or V2E. If no such validation is available, the claims should be explicitly scoped to 'efficient online event generation' rather than 'faithful event camera simulation.'
  3. [§5.3, Fig. 8 and Table 3] The performance evaluation reports ranges over resolutions and interpolation factors but does not give repeated-run statistics, standard deviations, or the number of trials for each configuration. Because the conclusions about online operation and the 'less than 400 MB' memory bound depend on these numbers, please report means and variances over multiple runs, describe the measurement procedure (warm-up, number of rendered frames, timing synchronization), and state the GPU clock/memory state where relevant.
minor comments (6)
  1. [Abstract] The phrase 'enables supports online multimodal event-camera simulation' appears to contain a typo; it should be 'enables online multimodal event-camera simulation.'
  2. [Table 2 and §5.1] The DAVIS346 resolution is given inconsistently as 346×260 in Figure 8 and Section 5.1 but as 260×346 in Table 2. Please standardize the orientation notation.
  3. [Table 3] The GenX320 row at ×32 reports 399.3 MB (grayscale) and 399.7 MB (Bayer). The claim 'less than 400 MB' is technically true but very close to the bound; consider reporting a margin or stating the measurement precision.
  4. [§4.2] The term 'backward warping' should be defined in one sentence, specifying the direction of the motion-vector field and the convention used by Isaac Sim, so readers can verify Eq. (4) independently.
  5. [Conclusions] The paper states that the add-on 'will be publicly released' but gives no repository URL, version, or installation details. Including a link would improve reproducibility and is important for a tool paper.
  6. [§4.2, Eqs. (1)--(3)] The per-pixel reference state update is described only as 'updated before processing the next sample.' Please clarify whether the reference is updated after every rendered and synthesized frame or after every event batch, since this affects the event-generation behavior when multiple intermediate samples are processed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: EsaacSim's event model is stated in-line from standard literature and the headline numbers are benchmark measurements, not predictions fitted to their own inputs.

full rationale

I walked the derivation chain of EsaacSim and found no step in which a claimed prediction reduces by construction to a fitted input or to an unverified self-citation. The event generation model in Section 4.2 (Equations 1–3) is a standard frame-based log-intensity contrast-threshold model, stated directly in the paper and consistent with the cited event-camera literature; it is not derived from, nor defined in terms of, the paper's own outputs. The Bayer RGGB extension is a straightforward application of the same threshold model to a virtual color filter array, and the claims about configurable thresholds, latency, and noise are parametric behaviors of the stated model, not fitted results later relabeled as predictions. The performance results in Section 5.3 (6.98–27.28 ms grayscale, 7.58–29.16 ms Bayer, under 400 MB GPU memory) are measured execution times and memory allocations; they are empirical benchmark data, so they cannot be circular in the sense of being forced by the model equations. The self-citation of Bugueno-Cordova et al. [4] appears in the related-work comparison and as inspiration for color event generation, but the load-bearing event model and the measured efficiency claims do not depend on that prior work as an unverified authority. The paper's own stated limitations in Section 5.3 — that the physics-grounded backend is not yet implemented and that no quantitative comparison against real event camera recordings has been performed — are acknowledged absence-of-validation statements rather than circular reasoning. The skeptic's concern about Equation (4) possibly being time-reversed relative to the timestamp assignment in Equation (5) is an internal correctness or consistency issue, not a circularity issue: even if the interpolation direction were wrong, the claimed event-generation behavior would not be equivalent to its inputs by construction. Applying the hard rule that circularity must be exhibited by a specific reduction, no such reduction exists here. Accordingly, the appropriate finding is no significant circularity, score 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no fitted free parameters; the event generation model is a standard threshold-on-log-intensity scheme from the cited literature. The load-bearing assumptions are: (1) the renderer's HDR images approximate sensor irradiance, (2) motion vectors support correct intermediate frame synthesis, and (3) the configured sensor parameters (threshold, noise, latency) are sufficient to emulate real event cameras. No new physical entities are posited.

assumptions (3)
  • domain assumption The frame-based log-intensity contrast threshold model (Equations 1-3) adequately emulates event camera behavior.
    Adopted from ESIM/V2E literature; the paper does not validate it against real event camera data (Section 5.3).
  • domain assumption Isaac Sim rendered HDR images and motion vectors provide accurate radiance and displacement information for event generation.
    The event backend consumes native RTX camera outputs and motion vectors (Sections 4.1, 4.2); if these are inaccurate, synthesized events are artifacts.
  • ad hoc to paper Backward warping of the current rendered image with renderer motion vectors (Equation 4) produces valid intermediate intensity frames.
    This is the paper's motion-guided frame-gap synthesis; it assumes motion is locally linear and the motion field is dense and correct, with fallback to linear interpolation when invalid (Section 4.2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of EsaacSim: A Multimodal Event Camera Add-on for NVIDIA Isaac Sim." pith.science (2026). https://pith.science/paper/C677N3LO

@misc{pith2026260808522,
  author       = {Pith},
  title        = {Pith review of: EsaacSim: A Multimodal Event Camera Add-on for NVIDIA Isaac Sim},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C677N3LO}},
  note         = {Machine review of arXiv:2608.08522}
}
read the original abstract

Event-based vision is becoming an increasingly important sensing paradigm for robotics, yet its adoption remains limited by sensor availability and the lack of integrated simulation tools for modern robotics platforms. This paper presents EsaacSim, a multimodal event camera add-on for NVIDIA Isaac Sim that enables online simulation of configurable event cameras with grayscale and Bayer RGGB event generation. The framework supports multiple event camera resolutions and provides synchronized RGB, APS, event, depth, and IMU outputs through native ROS2 interfaces. A motion-guided frame-gap synthesis strategy further increases the effective temporal resolution while preserving compatibility with the Isaac Sim rendering pipeline. Experimental evaluation demonstrates synchronized multimodal simulation across representative robotic scenes and efficient online performance over five event camera resolutions at effective event rates from 240 to 960Hz. Event stream generation requires 6.98--27.28ms for grayscale events and 7.58--29.16ms for Bayer RGGB events while using less than 400MB of additional GPU memory on an NVIDIA RTX~4060 GPU. These results show that EsaacSim enables supports online multimodal event-camera simulation for robotics research and synthetic data generation. We release an early version of the simulator and report its current architecture and performance.

Figures

Figures reproduced from arXiv: 2608.08522 by the authors.

Figure 1
Figure 1. Temporal evolution of the logarithmic intensity signal L(t) . Positive and negative events are generated whenever the accumulated logarithmic intensity variation reaches the corresponding contrast thresholds. Bayer RGGB event streams. In contrast to EVIS [32], which primarily provides a physics-grounded event camera plugin for RGB-event simulation through a Python API, EsaacSim is designed as a multimodal robotics a… view at source ↗
Figure 2
Figure 2. Layered architecture of EsaacSim within NVIDIA Isaac Sim. Native Isaac Sim components (green) provide scene simulation and camera render products (RGB, HDR, and depth). EsaacSim (red) acquires the RGB render products to generate APS frames and applies a configurable event camera model to produce either grayscale or Bayer RGGB event streams. RGB, APS, event, depth, and IMU data are synchronized and exposed through RO… view at source ↗
Figure 3
Figure 3. Runtime deployment of EsaacSim. The extension executes inside the Docker￾based Isaac Sim environment and exposes generated data through ROS 2 publishers and native output interfaces. ROS 2 applications run in the host environment managed with pixi. 4.2 Event Generation First, rendered images are converted into asynchronous events using a configurable log-intensity model. Subsequently, motion-guided frame-gap synthes… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Motion-vector-based frame-gap synthesis in EsaacSim. Given two consecutive rendered images, motion information is used to synthesize N − 1 intermediate intensity samples at uniformly spaced timestamps. Both rendered and synthesized samples are processed by the same fra…
Figure 5
Figure 5. Figure 5: ROS 2 message interfaces provided by EsaacSim for event-stream communica￾tion. Individual events are represented by Event, conventional batched communication uses EventPacket, and PackedEventPacket adopts a structure-of-arrays layout to re￾duce serialization overhead i…
Figure 6
Figure 6. Figure 6: Qualitative results produced by EsaacSim on representative YCB objects using a simulated DAVIS346 event camera (346 × 260 resolution). Each row corresponds to a different object, while columns show the synchronized RGB image, APS grayscale image, grayscale event visual…
Figure 7
Figure 7. Figure 7: Effect of the configurable contrast threshold on event generation. Left: total, positive, and negative event counts obtained from the same Power Drill sequence while varying the contrast threshold. Right: corresponding event visualizations for thresholds of 0.00, 0.10,…
Figure 8
Figure 8. Figure 8: shows the average event stream generation time per rendered-frame interval as a function of sensor resolution for both grayscale and Bayer RGGB event generation. The reported execution times correspond to the processing time required by the event-generation backend to …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 28 canonical work pages

  1. [1]

    Event camera - airsim (2021),https://microsoft.github.io/AirSim/event_sim/ 3, 4

    work page 2021

  2. [2]

    In: 2017 IEEE International Conference on Image Processing (ICIP)

    Bi, Y., Andreopoulos, Y.: Pix2nvs: Parameterized conversion of pixel-domain video frames to neuromorphic vision streams. In: 2017 IEEE International Conference on Image Processing (ICIP). pp. 1990–1994 (2017) 3

    work page 2017

  3. [3]

    The International Journal of Robotics Research45(2), 204–232 (2026) 2 EsaacSim 15

    Bonetto, E., Xu, C., Ahmad, A.: Grade: Generating realistic and dynamic environ- ments for robotics research with isaac sim. The International Journal of Robotics Research45(2), 204–232 (2026) 2 EsaacSim 15

    work page 2026

  4. [4]

    In: International Conference on Robotics, Computer Vision and Intelligent Systems

    Bugueno-Cordova, I., Campusano, M., Guaman-Rivera, R., Verschae, R.: A color event-based camera emulator for robot vision. In: International Conference on Robotics, Computer Vision and Intelligent Systems. pp. 375–390. Springer (2024) 3

    work page 2024

  5. [5]

    In: 2015 International Conference on Advanced Robotics (ICAR)

    Calli, B., Singh, A., Walsman, A., Srinivasa, S., Abbeel, P., Dollar, A.M.: The ycb object and model set: Towards common benchmarks for manipulation research. In: 2015 International Conference on Advanced Robotics (ICAR). pp. 510–517 (2015). https://doi.org/10.1109/ICAR.2015.725150410

    arXiv 2015

  6. [6]

    Information15(8) (2024)

    Cazzato, D., Bono, F.: An application-driven survey on event-based neuromor- phic computer vision. Information15(8) (2024). https://doi.org/10.3390/ info15080472,https://www.mdpi.com/2078-2489/15/8/4721

    work page 2024

  7. [7]

    In: European Conference on Computer Vision

    Chakravarthi, B., Verma, A.A., Daniilidis, K., Fermuller, C., Yang, Y.: Recent event camera innovations: A survey. In: European Conference on Computer Vision. pp. 342–376. Springer (2024) 1

    2024

  8. [8]

    Nature652(8111), 886–891 (2026).https://doi.org/10.1038/s41586-026-10338-52

    Dürr, P., El Gheche, M., Maeda, G.J., Mukai, N., Takahashi, N., Heusser, S., Sahloul, H., Saraiji, Y., Adodin, P., Bi, Y., Blakeman, S., Conti, C., Fuentes Hitos, D., Hu, Y., Khadivar, F., Kreiser, R., Martinez, L., Schilling, F., Tapiador Morales, R., Torrente, G., Ynocente Castro, M., Abecassis, L., Giammarino, A., Huang, Y.T., Nagel, Y., Scotti, A., Si...

    work page 2026

Show all 36 references
  1. [9]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 44(1), 154–180 (2022) 1, 4

    Gallego, G., Delbrück, T., Orchard, G., Bartolozzi, C., Taba, B., Censi, A., Leuteneg- ger, S., Davison, A.J., Conradt, J., Daniilidis, K., Scaramuzza, D.: Event-based vision: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 44(1), 154–180 (2022) 1, 4

    2022

  2. [10]

    arXiv preprint arXiv:2606.03551 (2026) 2

    Gao, S., Pagnucco, M., Bednarz, T., Song, Y.: Nvidia isaac sim: Enabling scalable, gpu-accelerated simulation for robotics. arXiv preprint arXiv:2606.03551 (2026) 2

    2026 arXiv

  3. [11]

    In: 2016 IEEE Sym- posium Series on Computational Intelligence (SSCI)

    García, G.P., Camilleri, P., Liu, Q., Furber, S.: pydvs: An extensible, real-time dynamic vision sensor emulator using off-the-shelf hardware. In: 2016 IEEE Sym- posium Series on Computational Intelligence (SSCI). pp. 1–7 (2016) 3

    2016

  4. [12]

    In: Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Tech- niques, and Applications III

    Greene, J.L., Kar, A., Galindo, I., Quiles, E., Chen, E., Anderson, M.: A pytorch- enabled tool for synthetic event camera data generation and algorithm development. In: Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Tech- niques, and Applications III....

    2025

  5. [13]

    Hageman, O.: A Survey on Event Camera Simulators and Datasets for Optical Flow Estimation. Ph.D. thesis, Delft University of Technology (2024) 2

    2024

  6. [14]

    In: European Conference on Computer Vision

    Han, H., Lyu, J., Li, J., Wei, H., Li, C., Wei, Y., Chen, S., Ji, X.: Physical-based event camera simulator. In: European Conference on Computer Vision. pp. 19–35. Springer (2024) 3, 4

    2024

  7. [15]

    In: 2025 IEEE International Conference on Real-time Computing and Robotics (RCAR)

    Hong, Y., Lu, T., Chen, J., Zhu, D., Wang, Z.: A comprehensive review of event camera-based depth estimation methods. In: 2025 IEEE International Conference on Real-time Computing and Robotics (RCAR). pp. 1046–1053 (2025).https: //doi.org/10.1109/RCAR65431.2025.111396131

    2025

  8. [16]

    In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

    Hu, Y., Liu, S.C., Delbruck, T.: v2e: From video frames to realistic dvs events. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). pp. 1312–1321 (2021) 2, 3

    2021

  9. [17]

    Jacinto, M., Pinto, J., Patrikar, J., Keller, J., Cunha, R., Scherer, S., Pascoal, A.: Pegasus simulator: An isaac sim framework for multiple aerial vehicles simulation. 16 I. Bugueno-Cordova et al. In: 2024 International Conference on Unmanned Aircraft Systems (ICUAS). pp. 91...

    2024

  10. [18]

    Frontiers in Neuroscience15, 702765 (2021) 3, 4

    Joubert, D., Marcireau, A., Ralph, N., Jolley, A., Van Schaik, A., Cohen, G.: Event camera simulator improvements via characterized parameters. Frontiers in Neuroscience15, 702765 (2021) 3, 4

    2021

  11. [19]

    In: 2016 IEEE International Conference on Simulation, Modeling, and Programming for Autonomous Robots

    Kaiser, J., Tieck, J.C.V., Hubschneider, C., Wolf, P., Weber, M., Hoff, M., Friedrich, A., Wojtasik, K., Roennau, A., Kohlhaas, R., Dillmann, R., Zöllner, J.M.: Towards a framework for end-to-end control of a simulated vehicle with spiking neural networks. In: 2016 IEEE Intern...

    2016

  12. [20]

    In: 2012 IEEE International Symposium on Circuits and Systems (ISCAS)

    Katz, M.L., Nikolic, K., Delbruck, T.: Live demonstration: Behavioural emulation of event-based vision sensors. In: 2012 IEEE International Symposium on Circuits and Systems (ISCAS). pp. 736–740 (2012) 3

    2012

  13. [21]

    In: British Machine Vision Conference (BMVC) (2018) 4

    Li, W., Saeedi, S., McCormac, J., Clark, R., Tzoumanikas, D., Ye, Q., Huang, Y., Tang, R., Leutenegger, S.: Interiornet: Mega-scale multi-sensor photo-realistic indoor scenes dataset. In: British Machine Vision Conference (BMVC) (2018) 4

    2018

  14. [22]

    Li, Z., Bai, X., Lu, J., Shen, P., Lam, E.Y., Peng, Y.: Eventtracer: Fast path tracing-based event stream rendering (2025) 2, 3, 4

    2025

  15. [23]

    The International Journal of Robotics Research36(2), 142–149 (2017) 4

    Mueggler, E., Rebecq, H., Gallego, G., Delbruck, T., Scaramuzza, D.: The event- camera dataset and simulator: Event-based data for pose estimation, visual odom- etry, and slam. The International Journal of Robotics Research36(2), 142–149 (2017) 4

    2017

  16. [24]

    In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

    Nehvi, J., Golyanik, V., Mueller, F., Seidel, H.P., Elgharib, M., Theobalt, C.: Differentiable event stream simulator for non-rigid 3d tracking. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). pp. 1302–1311 (2021) 4

    2021

  17. [25]

    Journal of Real-Time Image Processing19(2), 363–374 (2022) 4

    Pantho, M.J.H., Mbongue, J.M., Bhowmik, P., Bobda, C.: Event camera simulator design for modeling attention-based inference architectures. Journal of Real-Time Image Processing19(2), 363–374 (2022) 4

    2022

  18. [26]

    In: Billard, A., Dragan, A., Peters, J., Morimoto, J

    Rebecq, H., Gehrig, D., Scaramuzza, D.: Esim: an open event camera simulator. In: Billard, A., Dragan, A., Peters, J., Morimoto, J. (eds.) Proceedings of The 2nd Conference on Robot Learning. Proceedings of Machine Learning Research, vol. 87, pp. 969–982. PMLR (29–31 Oct 2018) 2, 3, 4

    2018

  19. [27]

    In: 2024 IEEE International Conference on Robotics and Automation (ICRA)

    Richard, A., Kamohara, J., Uno, K., Santra, S., van der Meer, D., Olivares-Mendez, M., Yoshida, K.: Omnilrs: A photorealistic simulator for lunar robotics. In: 2024 IEEE International Conference on Robotics and Automation (ICRA). pp. 16901– 16907 (2024).https://doi.org/10.1109...

    2024

  20. [28]

    In: Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP

    Rodríguez, D., Sonogashira, M., Kitano, K., Fujimura, Y., Funatomi, T., Mukaigawa, Y., Kawanishi, Y.: An event camera simulator for arbitrary viewpoints based on neural radiance fields. In: Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and ...

    2025 doi

  21. [29]

    In: 2021 International Conference on Communication information and Computing Technology (ICCICT)

    Shah, P., Rathod, S.S.: Review of bio-inspired silicon retina: From cell to system level implementation. In: 2021 International Conference on Communication information and Computing Technology (ICCICT). pp. 1–13. IEEE (2021) 4

    2021

  22. [30]

    In: Field and Service Robotics (2017) 4

    Shah, S., Dey, D., Lovett, C., Kapoor, A.: Airsim: High-fidelity visual and physical simulation for autonomous vehicles. In: Field and Service Robotics (2017) 4

    2017

  23. [31]

    IEEE Access12, 51275–51306 (2024).https://doi.org/10.1109/ACCESS.2024.33860321 EsaacSim 17

    Shariff, W., Dilmaghani, M.S., Kielty, P., Moustafa, M., Lemley, J., Corcoran, P.: Event cameras in automotive sensing: A review. IEEE Access12, 51275–51306 (2024).https://doi.org/10.1109/ACCESS.2024.33860321 EsaacSim 17

    2024

  24. [32]

    arXiv preprint arXiv:2607.08098 (2026) 2, 3, 4, 5

    Shi, L., Zhang, R., Wang, Z.: Evis: A physics-grounded event camera plugin for nvidia isaac sim. arXiv preprint arXiv:2607.08098 (2026) 2, 3, 4, 5

    2026 arXiv

  25. [33]

    Biomimetics9(7) (2024)

    Tenzin, S., Rassau, A., Chai, D.: Application of event cameras and neuromorphic computing to vslam: A survey. Biomimetics9(7) (2024). https://doi.org/10. 3390/biomimetics9070444,https://www.mdpi.com/2313-7673/9/7/4441

    2024

  26. [34]

    ACM Comput

    Wang, H., Guo, R., Ma, P., Ruan, C., Luo, X., Ding, W., Zhong, T., Xu, J., Liu, Y., Chen, X.: Event camera meets mobile embodied perception: Abstraction, algorithm, acceleration, application. ACM Comput. Surv.58(8) (Feb 2026).https: //doi.org/10.1145/3786332,https://doi.org/10...

    2026 doi

  27. [35]

    IEEE Transactions on Human-Machine Systems56(1), 32–47 (2026).https://doi

    Zafar, M.H., Moosavi, S.K.R., Sanfilippo, F.: Applications of neuromorphic/event camerainroboticswithhumaninloop:Asystematicreview,datasets,andchallenges. IEEE Transactions on Human-Machine Systems56(1), 32–47 (2026).https://doi. org/10.1109/THMS.2025.36280641, 2

    2026

  28. [36]

    In: 2023 IEEE International Conference on Robotics and Automation (ICRA)

    Ziegler, A., Teigland, D., Tebbe, J., Gossard, T., Zell, A.: Real-time event simulation with frame-based cameras. In: 2023 IEEE International Conference on Robotics and Automation (ICRA). pp. 11669–11675 (May 2023).https://doi.org/10.1109/ ICRA48891.2023.101606543

    2023

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.