SNM-VFI: Symmetric Nonlinear Motion-Guided Generative Video Frame Interpolation

Fatih Porikli; Hanno Ackermann; Hong Cai; Hyeonjun Sim; Jamie Menjay Lin; Jisoo Jeong; Yinhao Zhu; Yunxiao Shi

REVIEW 4 major objections 6 minor 34 references

SNM-VFI: Symmetric Nonlinear Motion-Guided Generative Video Frame Interpolation

T0 review · 4 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 Training-free video frame interpolation: flow-guided latents steer a pre-trained video diffusion model to stronger perceptual quality.

desk verdict Solid training-free VFI pipeline whose real effect is in the latent blending, but the headline benchmark tables compare four-frame inputs against two-frame baselines. read the letter →

arxiv 2608.13460 v1 pith:VZJ36ZA7 submitted 2026-08-13 cs.CV

Jisoo Jeong , Hong Cai , Jamie Menjay Lin , Hanno Ackermann , Hyeonjun Sim , Yinhao Zhu , Yunxiao Shi , Fatih Porikli This is my paper · ORCID

classification cs.CV

keywords videoframeinterpolationopticalflowdiffusionmodeltraining-freemotionguidancelatentpriortemporalcoherenceconfidence-awarefusion

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 claims that state-of-the-art video frame interpolation can be achieved without training any task-specific network, by combining pre-trained optical flow with a pre-trained video diffusion model. The flow model produces a nonlinear-motion intermediate frame that, once encoded into the diffusion model's latent space, replaces random noise and is re-blended into the denoising trajectory at every step. This preserves dense motion correspondence while letting the diffusion model generate realistic texture, and a confidence map then fuses the flow frame's reliable structure with diffusion-generated details in occluded and boundary regions. On DAVIS, Sintel, and KITTI, the method reports the best LPIPS scores and competitive PSNR/SSIM and FID across both x2 and x4 interpolation, including gains over the diffusion-only baseline. The wider significance is that explicit motion priors can adapt off-the-shelf generative video models to low-level temporal tasks without fine-tuning.

What carries the argument

The load-bearing mechanism is the flow-guided latent prior: the latent vector z^K = VAE_e(I_hat_t) obtained by encoding the flow-warped intermediate frame, combined with the iterative blending rule z^(k-1) <- (k/K)*z^K + (1 - k/K)*z^(k-1)_diff, where k decreases from K to 0. Because the weight is k/K, the flow prior dominates early denoising steps and the diffusion output takes over later, letting explicit motion correspondence steer the generative process without fighting the learned latent distribution. The second mechanism is the symmetric nonlinear flow model of Eq. 4, which uses neighbor flows V_(0->-1) and V_(1->2), applied symmetrically at both input frames, to approximate the intermediate flow V_(0->t), together with the occlusion map M of Eq. 5 that reverts uncertain regions to stable linear motion. The third mechanism is the confidence-aware fusion I_t = C_t*I_t^Flow + (1 - C_t)*I_t^Diff of Eq. 9, which keeps flow-based structure where forward-backward consistency is high and lets diffusion fill occlusions and boundaries.

What would settle it

Run the exact SNM-VFI pipeline but replace the flow-guided latent z^K with (a) random Gaussian noise and (b) the latent of an incorrectly warped frame, keeping all other steps identical; if DAVIS LPIPS and FID do not clearly worsen for both replacements, the claimed role of the flow prior is not supported. Conversely, swapping the pre-trained video diffusion model (and its VAE) for a different one and observing whether the same gains persist would test whether the latent-blending mechanism is generic or specific to GenIn's fine-tuned weights.

Watch

Extended reading notes

Core claim

SNM-VFI's central claim is that a pre-trained latent video diffusion model can be made temporally coherent and perceptually strong at frame interpolation by (1) initializing its latent trajectory with the VAE encoding of a flow-based intermediate frame and (2) blending that flow-guided latent back into the denoised latent at every step with a weight that decays linearly from full to nearly zero. The discovery is that this latent-level motion guidance, rather than input conditioning or fine-tuning, is what lets the diffusion model preserve object trajectories and avoid the brightness and contrast drift and static-object artifacts typical of diffusion-only interpolation. The paper further claims that the flow frame itself must come from a symmetric nonlinear motion model using past and future neighbor flows applied at both endpoints, with an occlusion-aware fallback to linear motion, and that a confidence-aware fusion of the flow and diffusion outputs yields the final perceptual gain. These claims are supported by ablations showing each component (symmetric nonlinear motion, occlusion handling, latent initialization, iterative latent blending, confidence fusion) improves the aggregate on the three benchmarks.

Load-bearing premise

The pipeline assumes that the VAE-encoded flow-warped frame lies close enough to the learned latent distribution of the pre-trained video diffusion model that replacing random noise with it and blending it back in during denoising will stay on the generative manifold; if that fails, the diffusion output becomes unstable or artifact-prone and the perceptual gains over GenIn would disappear.

Editorial extensions

If this is right

Any pre-trained optical flow model and any pre-trained latent video diffusion model can be combined with this recipe, since no weights are updated and only a VAE encoder and decoder are used as-is.
The same flow-guided latent prior extends from x2 to x4 interpolation, synthesizing I_1/4, I_2/4, and I_3/4 with consistent gains over both flow-only and diffusion-only baselines.
The strong initialization from the flow frame reduces the required number of diffusion steps from 50 to 20, cutting compute roughly by a factor of 2.5 while keeping quality.
Confidence-aware fusion lets the framework keep the structural precision of flow in well-matched regions while using diffusion-generated detail exactly where correspondences fail, such as occlusions and object boundaries.
Temporal coherence is robust across datasets with very different motion statistics, from surgical DAVIS sequences to fast KITTI driving scenes, as reflected by the best LPIPS and near-best FID numbers.

Reading between the lines

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

The same latent-guidance idea should transfer to other generation tasks where a classical or geometric prior exists, such as video deblurring, super-resolution, or novel view synthesis from sparse views, replacing fine-tuning with latent-space steering.
Because the paper uses four input frames (the pair plus one past and one future neighbor), a direct testable extension is to measure how performance degrades when only the two frame endpoints are available but the same symmetric flow formula is used with the absent neighbor flows set to zero.
The improvement from blending latents rather than conditioning on the flow frame suggests that the diffusion latent space is a sufficiently expressive host for explicit motion constraints; this could be probed directly by measuring the distance between z^K and the diffusion model's learned trajectory under the same conditioning.

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

Quick take: this is a competent training-free pipeline that injects optical-flow priors into a pretrained video diffusion model, and the best evidence is its own ablation: adding the per-step latent blending (Eq. 7) to GenIn cuts LPIPS on DAVIS from 0.297 to 0.234. The mechanism is plausible. But the headline comparisons overstate the gains because SNM-VFI uses four input frames (I−2, I0, I2, I4) while every baseline gets two (I0, I2). The paper discloses this only in Section 7.2 and presents Table 1 as an apples-to-apples state-of-the-art comparison. That is the main soft spot.

What's actually new: the symmetric nonlinear flow interpolation in Eq. 4 with occlusion handling, and the flow-guided latent initialization plus per-step blending in Eq. 7. Individual pieces come from OCAI, GenIn, and quadratic VFI, but the combination, plus the confidence-aware fusion of Eq. 9, is this paper's contribution. Credit where it's due: the derivations are coherent, the ablations are fairly complete, all components use public pretrained weights, the self-citation to OCAI is transparent, and the KITTI ×2 numbers are genuinely good across all four metrics.

The soft spots are real but fixable. First, no two-frame variant of SNM-VFI is reported, so we cannot tell how much of the win comes from the extra frames. Second, alpha = 0.5 is selected on the same three benchmark test sets, with no held-out split; the differences are small, but benchmark tuning cannot be excluded. Third, your stress-test point is correct: flow init alone degrades LPIPS/FID on DAVIS and KITTI in Table 3, so the perceptual gains come from the blending, not the init. The paper never isolates blending without init, nor does it check whether the VAE-encoded flow frame stays near the diffusion model's expected latent manifold. That does not sink the method — init plus blending is the intended design and it works — but the claim that flow priors 'preserve dense motion correspondence' needs more support. Fourth, no code and no error bars.

Who is this for: people building diffusion-based VFI systems, and anyone studying how to condition video diffusion with explicit motion cues. It deserves peer review — the idea is sound, the ablation is useful, and the confounds are addressable. I'd recommend sending it out, asking for a two-frame variant, a held-out alpha choice, and an ablation that separates init from blending. With those, it would be a solid accept.

Referee Report

4 major / 6 minor

Summary. The paper proposes SNM-VFI, a training-free video frame interpolation framework that combines a pre-trained optical flow model (RAFT) with a pre-trained video diffusion model (GenIn/SVD) to generate intermediate frames. The method first constructs flow-based intermediate frames and confidence maps via a symmetric nonlinear motion model that uses four input frames, then encodes these frames as latent priors to initialize and iteratively guide the diffusion denoising process (Eq. 7), and finally fuses the flow-based and diffusion-based outputs using confidence maps (Eq. 9). The authors evaluate on DAVIS, Sintel, and KITTI under x2 and x4 interpolation settings, reporting PSNR, SSIM, LPIPS, and FID, and claim strong perceptual quality, competitive reconstruction accuracy, and robust temporal coherence.

Significance. If the results hold, the paper would demonstrate a useful training-free recipe for adapting off-the-shelf video diffusion models to low-level temporal interpolation, avoiding task-specific fine-tuning. The paper is commendably transparent about its use of multiple input frames, reports ablations for each design component, and uses official implementations and pretrained weights for all baselines. However, the main quantitative comparisons suffer from an input-frame asymmetry in which SNM-VFI sees four frames while most baselines see only two, and the closest multi-frame baselines are confined to the ablation table. The latent-manifold premise underlying the diffusion guidance is also only weakly supported, and the alpha hyperparameter is selected on the evaluation benchmarks. These issues mean the headline claims are not yet fully supported, though they are plausibly addressable with additional controlled experiments.

major comments (4)

[Section 4.1, Tables 1-2] The main quantitative comparisons are not controlled for the number of input frames. Section 4.1 explicitly states that SNM-VFI uses four frames (I-2, I0, I2, I4) in the x2 setting, while 'most existing algorithms' use only two frames (I0, I2). The closest four-frame flow-based baselines, Quadratic [30] and Enhanced Quadratic [17], appear only in the ablation study (Table 3) and only for the x4 setting, not in Tables 1-2 where the headline claims are made. This asymmetry alone could explain part of the reported gains, especially for reconstruction-oriented metrics. I ask the authors to add the multi-frame baselines (Quadratic and Enhanced Quadratic) to Tables 1 and 2, or alternatively to include a two-frame variant of SNM-VFI, so that the comparison is meaningful.
[Table 3, diffusion ablation rows] The ablation evidence undercuts the paper's central latent-trajectory premise. In Table 3, 'Baseline + flow init' (using z^K only as initial noise, without the blending of Eq. 7) degrades LPIPS and FID relative to the GenIn baseline on DAVIS (LPIPS 0.2967 to 0.3152, FID 54.99 to 62.76) and on KITTI (FID 25.69 to 28.13). The gains appear only after adding the weighted latent blending of Eq. 7 and the confidence fusion of Eq. 9. The paper never verifies that z^K lies on or near the latent distribution expected by the pretrained diffusion model, nor does it provide a reason why the linear schedule w_k = k/K should keep the process on-manifold. I recommend an explicit test of the manifold assumption, for example by comparing latent statistics or denoising trajectories between random-init and flow-init runs, and an ablation that separates the effect of Eq. 7 from the effect of Eq. 9 more cleanly than the current rows do.
[Section 5, alpha hyperparameter, Table 3] The nonlinearity strength alpha in Eq. 4 is selected on the same benchmark datasets used for the final evaluation. The second section of Table 3 optimizes alpha over {1/4, 2/4, 3/4, 1.0} on DAVIS, Sintel, and KITTI, and the paper then adopts alpha = 0.5 for all reported results in Tables 1 and 2. Since SNM-VFI is training-free, this benchmark-driven selection of its only tunable scalar is a real source of optimistic bias. I ask the authors to either choose alpha on a held-out validation split or report a sensitivity analysis that clearly shows the final conclusions are invariant to alpha across the full range.
[Abstract and Section 4.2] The claim of 'robust temporal coherence' is not supported by any quantitative temporal-consistency metric. Tables 1 and 2 report per-frame image quality measures (PSNR, SSIM, LPIPS, FID) with each frame evaluated independently; there is no video-level or trajectory-level measure such as warped-frame error, temporal flicker, or flow-consistency error between adjacent interpolated frames. The qualitative examples in Figures 4-10 are suggestive, but the abstract's explicit claim about temporal coherence needs either direct measurement or a careful softening of the wording.

minor comments (6)

[Section 7.1] The sentence about noise reinjection is unclear: 'our method applies noise reinjection once at each steps throughout the entire diffusion process' appears grammatically broken and the intended schedule (whether noise is reinjected at every step, once, or not at all) is not described.
[Section 7.1] The text lists temporal positions as 't=1/24, 2/24, 3/34, ..., 23/24'; the entry '3/34' should presumably be '3/24'.
[Section 7.2] There is a typo: 'datsets' should be 'datasets'.
[Figures 6-10] Several figure captions refer to 'EVA-VFI' when the method in question is EMA-VFI; please correct the captions and any in-text references.
[Section 3.2, Eq. 7] The notation for the flow-guided latent is unclear: Section 3.2 defines z^K_n for a set of intermediate frames, but Eq. 7 uses a single z^K without explaining how per-frame latents are indexed or aggregated during the video-diffusion process.
[Section 5, linear model FID discussion] The speculation that the linear model's better FID arises because it 'may generate smoother and more simplified intermediate frames' is not tested; either remove or support this hypothesis.

Circularity Check

0 steps flagged · score 0.0 of 10

No definitional circularity: the flow-guided diffusion recipe (Eq. 7) and confidence fusion (Eq. 9) are constructive and are verified against external baselines, not derived from their own inputs.

full rationale

The paper's central mechanism is a concrete recipe: flow-based intermediate frames and confidence maps are computed by Eqs. 4-6 from pretrained RAFT optical flow, encoded by a VAE, and then injected into a pretrained GenIn/SVD diffusion model through the explicit weighted blending schedule in Eq. 7; the final output is a confidence-weighted combination of the flow and diffusion outputs in Eq. 9. Nothing in these equations is defined in terms of the target benchmark scores or the final output, and the claimed gains are established by comparisons against external baselines (IFRNet, VFIFormer, AMT, EMA-VFI, BiM-VFI, RIPR, OCAI, LDMVFI, TRF, GenIn) using public pretrained weights. The self-citations to OCAI [13] and DistractFlow [14] supply building blocks (flow decomposition, hole masks, forward-backward confidence), but these are transparent, externally published components rather than an imported uniqueness theorem or an unverified load-bearing premise. The ablation in Table 3 shows that flow initialization alone can degrade LPIPS/FID relative to GenIn, so the reported improvements are not a tautology of the design; they depend on the specific weighting and fusion. Concerns such as the benchmark-selected alpha=0.5, the reduced 20-step diffusion budget, and the four-frame versus two-frame input asymmetry in the main tables are validity and fairness issues, not instances of the paper's predictions reducing to its inputs by construction.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central pipeline relies on pretrained optical flow and video diffusion components and on one tuned hyperparameter; it does not introduce new physical entities. All assumptions are empirical and technical rather than mathematical postulates.

free parameters (1)

alpha (nonlinearity strength) = 0.5
Controls the magnitude of the symmetric nonlinear correction in Eq. 4; selected after ablating alpha = 1/4, 2/4, 3/4, 1.0 on the DAVIS/Sintel/KITTI x4 benchmarks (Table 3) and then applied to all experiments.

assumptions (5)

standard math The optical-flow decomposition identity V0->1(x) = V0->t(x) + Vt->1(x + V0->t(x)) from OCAI holds with pre-trained RAFT flows.
Adopted from OCAI Eq. 1 and used in Sec. 3.1 to derive Vt->1; it follows from flow composition if the flows are locally consistent.
domain assumption Pre-trained RAFT provides sufficiently accurate V0->-1 and V1->2 flows for the symmetric nonlinear correction in Eq. 4.
The whole motion prior depends on these extra flows being accurate enough to improve, not degrade, interpolation.
domain assumption The occlusion and matching maps from UnFlow (Eq. 5) correctly identify regions where nonlinear flow should fall back to a linear approximation.
Eq. 5 replaces flow with the OCAI linear flow where the matching map is 1; if occlusion detection fails, warping errors persist.
domain assumption A VAE-encoded flow-based frame z^K is a valid initial latent for the pretrained GenIn/SVD diffusion model, and per-step blending in Eq. 7 keeps samples on-manifold.
Eq. 7 assumes the flow latent and the denoised latents can be linearly combined without out-of-distribution artifacts.
domain assumption Forward-backward flow consistency yields confidence maps that are reliable for the fusion in Eq. 9.
Confidence weighting determines how much of the flow-based frame is kept; if the confidence is miscalibrated, fusion can propagate artifacts.

how reviews work

0 comments

Cite this review

Pith. "Pith review of SNM-VFI: Symmetric Nonlinear Motion-Guided Generative Video Frame Interpolation." pith.science (2026). https://pith.science/paper/VZJ36ZA7

@misc{pith2026260813460,
  author       = {Pith},
  title        = {Pith review of: SNM-VFI: Symmetric Nonlinear Motion-Guided Generative Video Frame Interpolation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VZJ36ZA7}},
  note         = {Machine review of arXiv:2608.13460}
}

read the original abstract

We propose Symmetric Nonlinear Motion-guided Generative Video Frame Interpolation (SNM-VFI), a training-free framework for motion-controllable generative video frame interpolation with pre-trained optical flow and video diffusion models. Unlike conventional diffusion-based VFI methods that synthesize intermediate frames from random noise, SNM-VFI guides the generative process with correspondence-aware frames produced by a symmetric nonlinear motion model. Specifically, we first utilize a pre-trained optical flow model to construct multi-frame nonlinear flow-based intermediate frames and confidence maps. These flow-guided frames are then encoded as latent priors to initialize and iteratively guide a pre-trained Video Diffusion model, enabling the diffusion model to preserve dense motion correspondence while improving perceptual realism. To further enhance output quality, we employ confidence maps to fuse structurally reliable flow-based predictions with diffusion-generated details in uncertain regions such as occlusions and object boundaries. Extensive evaluations on challenging benchmarks, including DAVIS, Sintel, and KITTI, demonstrate that SNM-VFI achieves strong perceptual quality, competitive reconstruction accuracy, and robust temporal coherence across diverse motion scenarios.

Figures

Figures reproduced from arXiv: 2608.13460 by the authors.

**Figure 1.** Video Frame Interpolation results (Iˆ 1/4, Iˆ 2/4, Iˆ 3/4) on the KITTI dataset. VFIFormer [18] shows interpolation results produced using a linear flow-based method, while LDMVFI [3] and GenIn [27] present results from diffusion-based VFI approaches. Quadratic [30] represents a multi-frame VFI method, and the final column displays the results generated by our approach. In contrast, recent advances in generative mo… view at source ↗

**Figure 2.** Linear motion (left), Quadratic Motion (middle), and Symmetric Nonlinear Motion (right). Linear motion assumes constant velocity, causing the optical flows (V0→t and V1→t) to meet at a single point, but failing to capture complex real-world dynamics. Quadratic motion incorporates velocity and acceleration at I0 and I1 by leveraging additional optical flows (V0→−1, V1→2), yet forward and backward flows do not always … view at source ↗

**Figure 3.** Overview of SNM-VFI as a motion-guided generative visual processing framework. SNM-VFI first constructs correspondence-aware intermediate frames and confidence maps using symmetric nonlinear motion priors (Sec. 3.1). These flow-guided frames are encoded as latent priors to initialize and iteratively guide a pre-trained video diffusion model, producing temporally coherent generative interpolations (Sec. 3.2). Final… view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Video frame interpolation results (×2 interpolation) on DAVIS, Sintel, and KITTI. The fourth to sixth rows show results from flow-based methods: VFIFormer and EMA-VFI (backward warping) and OCAI (forward warping). The seventh and eighth rows present results from diffus…

**Figure 5.** Figure 5: Video frame interpolation results (×4 interpolation) on KITTI. The second and third columns show results from flow-based methods: VFIFormer (backward warping) and OCAI (forward warping). The fourth column presents result from diffusion-based method: GenIn. The final co…

**Figure 6.** Figure 6: Video frame interpolation results on KITTI (×2). The second to fifth rows show results from flow-based methods: VFIFormer and EVA-VFI (backward warping) and OCAI (forward warping). The sixth and seventh rows present results from diffusionbased methods: LDMVFI and GenI…

**Figure 7.** Figure 7: Video frame interpolation results on Sintel (×2). The second to fifth rows show results from flow-based methods: VFIFormer and EVA-VFI (backward warping) and OCAI (forward warping). The sixth and seventh rows present results from diffusionbased methods: LDMVFI and Gen…

**Figure 8.** Figure 8: Video frame interpolation results on DAVIS (×2). The second to fifth rows show results from flow-based methods: VFIFormer and EVA-VFI (backward warping) and OCAI (forward warping). The sixth and seventh rows present results from diffusionbased methods: LDMVFI and GenI…

**Figure 9.** Figure 9: Video frame interpolation results on KITTI (×4). The second to fourth rows show results from flow-based methods: EVA-VFI and Bim-VFI (backward warping) and OCAI (forward warping). The fifth and sixth rows present results from diffusionbased methods: LDMVFI and GenIn. …

**Figure 10.** Figure 10: Video frame interpolation results on DAVIS (×4). The second to fourth rows show results from flow-based methods: EVA-VFI and Bim-VFI (backward warping) and OCAI (forward warping). The fifth and sixth rows present results from diffusionbased methods: LDMVFI and GenIn.…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 22 canonical work pages

[30]

Advances in Neural Information Processing Systems32(2019)

Xu, X., Siyao, L., Sun, W., Yin, Q., Yang, M.H.: Quadratic video interpolation. Advances in Neural Information Processing Systems32(2019)

work page 2019
[17]

In: European conference on computer vision

Liu, Y., Xie, L., Siyao, L., Sun, W., Qiao, Y., Dong, C.: Enhanced quadratic video interpolation. In: European conference on computer vision. pp. 41–56. Springer (2020)

work page 2020
[1]

arXiv preprint arXiv:2311.15127 (2023)

Blattmann, A., Dockhorn, T., Kulal, S., Mendelevitch, D., Kilian, M., Lorenz, D., Levi, Y., English, Z., Voleti, V., Letts, A., et al.: Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127 (2023)

arXiv 2023
[2]

In: Proceedings of the European Conference on Com- puter Vision

Butler, D.J., Wulff, J., Stanley, G.B., Black, M.J.: A naturalistic open source movie for optical flow evaluation. In: Proceedings of the European Conference on Com- puter Vision. pp. 611–625. Springer (2012)

work page 2012
[3]

In: Proceedings of the AAAI Conference on Artificial Intelligence

Danier, D., Zhang, F., Bull, D.: Ldmvfi: Video frame interpolation with latent diffusion models. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38, pp. 1472–1480 (2024)

2024
[4]

In: Proceedings of the IEEE/CVF International Conference on Computer Vision

Dosovitskiy, A., Fischer, P., Ilg, E., Hausser, P., Hazirbas, C., Golkov, V., Van Der Smagt, P., Cremers, D., Brox, T.: Flownet: Learning optical flow with convo- lutional networks. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2758–2766 (2015)

work page 2015
[5]

Feng, H., Ding, Z., Xia, Z., Niklaus, S., Abrevaya, V., Black, M.J., Zhang, X.: Explorativeinbetweeningoftimeandspace.In:EuropeanConferenceonComputer Vision. pp. 378–395. Springer (2024)

work page 2024
[6]

The International Journal of Robotics Research32(11), 1231–1237 (2013)

Geiger, A., Lenz, P., Stiller, C., Urtasun, R.: Vision meets robotics: The kitti dataset. The International Journal of Robotics Research32(11), 1231–1237 (2013)

2013

Show all 34 references

[7]

arXiv preprint arXiv:2207.11075 (2022)

Han, Y., Luo, K., Luo, A., Liu, J., Fan, H., Luo, G., Liu, S.: Realflow: Em-based re- alistic optical flow dataset generation from videos. arXiv preprint arXiv:2207.11075 (2022)

2022 arXiv
[8]

Advances in neural information processing systems30(2017)

Heusel,M.,Ramsauer,H.,Unterthiner,T.,Nessler,B.,Hochreiter,S.:Ganstrained by a two time-scale update rule converge to a local nash equilibrium. Advances in neural information processing systems30(2017)

2017
[9]

In: Proceedings of the European Conference on Computer Vision (2022)

Huang, Z., Shi, X., Zhang, C., Wang, Q., Cheung, K.C., Qin, H., Dai, J., Li, H.: Flowformer: A transformer architecture for optical flow. In: Proceedings of the European Conference on Computer Vision (2022)

2022
[10]

In: European Conference on Computer Vision

Huang, Z., Zhang, T., Heng, W., Shi, B., Zhou, S.: Real-time intermediate flow estimation for video frame interpolation. In: European Conference on Computer Vision. pp. 624–642. Springer (2022)

2022
[11]

arXiv preprint arXiv:1903.07414 (2019)

Hui, T.W., Tang, X., Loy, C.C.: A lightweight optical flow cnn-revisiting data fidelity and regularization. arXiv preprint arXiv:1903.07414 (2019)

2019 arXiv
[12]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Jain, S., Watson, D., Tabellion, E., Poole, B., Kontkanen, J., et al.: Video inter- polation with diffusion models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 7341–7351 (2024)

2024
[13]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Jeong, J., Cai, H., Garrepalli, R., Lin, J.M., Hayat, M., Porikli, F.: Ocai: Im- proving optical flow estimation by occlusion and consistency aware interpolation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 19352–19362 (2024)

2024
[14]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Jeong, J., Cai, H., Garrepalli, R., Porikli, F.: Distractflow: Improving optical flow estimation via realistic distractions and pseudo-labeling. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 13691– 13700 (2023)

2023
[15]

In: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion

Kong, L., Jiang, B., Luo, D., Chu, W., Huang, X., Tai, Y., Wang, C., Yang, J.: Ifrnet: Intermediate feature refine network for efficient frame interpolation. In: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion. pp. 1969–1978 (2022) 16 J. J...

2022
[16]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Li, Z., Zhu, Z.L., Han, L.H., Hou, Q., Guo, C.L., Cheng, M.M.: Amt: All-pairs multi-field transforms for efficient frame interpolation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 9801– 9810 (2023)

2023
[18]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Lu, L., Wu, R., Lin, H., Lu, J., Jia, J.: Video frame interpolation with transformer. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 3532–3542 (2022)

2022
[19]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Mayer, N., Ilg, E., Hausser, P., Fischer, P., Cremers, D., Dosovitskiy, A., Brox, T.: A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. ...

2016
[20]

In: Proceedings of the AAAI conference on artificial intelligence

Meister, S., Hur, J., Roth, S.: Unflow: Unsupervised learning of optical flow with a bidirectional census loss. In: Proceedings of the AAAI conference on artificial intelligence. vol. 32 (2018)

2018
[21]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Menze, M., Geiger, A.: Object scene flow for autonomous vehicles. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 3061–3070 (2015)

2015
[22]

In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Niklaus, S., Liu, F.: Softmax splatting for video frame interpolation. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 5437–5446 (2020)

2020
[23]

arXiv preprint arXiv:1704.00675 (2017)

Pont-Tuset, J., Perazzi, F., Caelles, S., Arbeláez, P., Sorkine-Hornung, A., Van Gool, L.: The 2017 davis challenge on video object segmentation. arXiv preprint arXiv:1704.00675 (2017)

2017 arXiv
[24]

In: Proceedings of the Computer Vision and Pattern Recognition Conference

Seo, W., Oh, J., Kim, M.: Bim-vfi: Bidirectional motion field-guided frame interpo- lation for video with non-uniform motions. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 7244–7253 (2025)

2025
[25]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Shang, W., Ren, D., Yang, Y., Zhang, H., Ma, K., Zuo, W.: Joint video multi- frame interpolation and deblurring under unknown exposure time. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 13935–13944 (2023)

2023
[26]

In: Proceedings of the European Conference on Computer Vision

Teed, Z., Deng, J.: Raft: Recurrent all-pairs field transforms for optical flow. In: Proceedings of the European Conference on Computer Vision. pp. 402–419. Springer (2020)

2020
[27]

arXiv preprint arXiv:2408.15239 (2024)

Wang, X., Zhou, B., Curless, B., Kemelmacher-Shlizerman, I., Holynski, A., Seitz, S.M.: Generative inbetweening: Adapting image-to-video models for keyframe in- terpolation. arXiv preprint arXiv:2408.15239 (2024)

2024 arXiv
[28]

IEEE transactions on image processing 13(4), 600–612 (2004)

Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing 13(4), 600–612 (2004)

2004
[29]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Xu, H., Zhang, J., Cai, J., Rezatofighi, H., Tao, D.: Gmflow: Learning optical flow via global matching. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 8121–8130 (2022)

2022
[31]

International Journal of Computer Vision127, 1106–1125 (2019)

Xue, T., Chen, B., Wu, J., Wei, D., Freeman, W.T.: Video enhancement with task- oriented flow. International Journal of Computer Vision127, 1106–1125 (2019)

2019
[32]

arXiv preprint arXiv:2501.03699 (2025) SNM-VFI: Motion-Guided Generative VFI 17

Zhang, G., Zhu, Y., Cui, Y., Zhao, X., Ma, K., Wang, L.: Motion-aware generative frame interpolation. arXiv preprint arXiv:2501.03699 (2025) SNM-VFI: Motion-Guided Generative VFI 17

2025 arXiv
[33]

In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Zhang, G., Zhu, Y., Wang, H., Chen, Y., Wu, G., Wang, L.: Extracting motion and appearance via inter-frame attention for efficient video frame interpolation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 5682–5692 (2023)

2023
[34]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The unreasonable effectiveness of deep features as a perceptual metric. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 586–595 (2018) 18 J. Jeong et al. 7 Implementation Details...

2018

Pith tools

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

[1] [30]

Advances in Neural Information Processing Systems32(2019)

Xu, X., Siyao, L., Sun, W., Yin, Q., Yang, M.H.: Quadratic video interpolation. Advances in Neural Information Processing Systems32(2019)

work page 2019

[2] [17]

In: European conference on computer vision

Liu, Y., Xie, L., Siyao, L., Sun, W., Qiao, Y., Dong, C.: Enhanced quadratic video interpolation. In: European conference on computer vision. pp. 41–56. Springer (2020)

work page 2020

[3] [1]

arXiv preprint arXiv:2311.15127 (2023)

Blattmann, A., Dockhorn, T., Kulal, S., Mendelevitch, D., Kilian, M., Lorenz, D., Levi, Y., English, Z., Voleti, V., Letts, A., et al.: Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127 (2023)

arXiv 2023

[4] [2]

In: Proceedings of the European Conference on Com- puter Vision

Butler, D.J., Wulff, J., Stanley, G.B., Black, M.J.: A naturalistic open source movie for optical flow evaluation. In: Proceedings of the European Conference on Com- puter Vision. pp. 611–625. Springer (2012)

work page 2012

[5] [3]

In: Proceedings of the AAAI Conference on Artificial Intelligence

Danier, D., Zhang, F., Bull, D.: Ldmvfi: Video frame interpolation with latent diffusion models. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38, pp. 1472–1480 (2024)

2024

[6] [4]

In: Proceedings of the IEEE/CVF International Conference on Computer Vision

Dosovitskiy, A., Fischer, P., Ilg, E., Hausser, P., Hazirbas, C., Golkov, V., Van Der Smagt, P., Cremers, D., Brox, T.: Flownet: Learning optical flow with convo- lutional networks. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2758–2766 (2015)

work page 2015

[7] [5]

Feng, H., Ding, Z., Xia, Z., Niklaus, S., Abrevaya, V., Black, M.J., Zhang, X.: Explorativeinbetweeningoftimeandspace.In:EuropeanConferenceonComputer Vision. pp. 378–395. Springer (2024)

work page 2024

[8] [6]

The International Journal of Robotics Research32(11), 1231–1237 (2013)

Geiger, A., Lenz, P., Stiller, C., Urtasun, R.: Vision meets robotics: The kitti dataset. The International Journal of Robotics Research32(11), 1231–1237 (2013)

2013