Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2603.26835

ANVIL: Accelerator-Native Video Interpolation via Codec Motion Vector Priors

Shibo Liu
View original source ↗Read source document ↗Export metadata ↓

Abstract

Real-time 30-to-60 fps video frame interpolation on mobile neural processing units (NPUs) requires each synthesized frame within 33.3 ms. We show that mainstream flow-based video frame interpolation faces three structural deployment barriers on mobile NPUs: spatial sampling operators exceed the frame budget or lack hardware support, iterative flow refinement collapses under 8-bit integer post-training quantization, and memory-bound operators dominate the inference graph. ANVIL addresses these barriers by reusing motion vectors from the H.264/AVC decoder to prealign input frames, removing learned optical flow, spatial sampling, and iterative accumulation from the accelerator graph. The remaining residual is refined by a convolution-dominated network composed almost entirely of compute-bound operators. On a Snapdragon 8 Gen 3 device, ANVIL achieves 12.8 ms 1080p inference at 8-bit integer precision; an open-source Android player sustains 28.4 ms median end-to-end latency over 30-minute continuous playback. Per-operator causal analysis identifies quantized accumulation on recurrent flow states as a key mechanism behind integer quantization failure in iterative methods. The current design targets H.264/AVC playback with decoder-exposed motion vectors.

Explore related subjects

eess.IVcs.CV

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Shibo Liu. 2026-04-01. ANVIL: Accelerator-Native Video Interpolation via Codec Motion Vector Priors. https://arxiv.org/abs/2603.26835

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Dipole-lets: a new multiscale decomposition for MR phase and quantitative susceptibility mapping

Nondipolar phase contributions can generate severe streaking artifacts during quantitative susceptibility mapping (QSM) inversion. We propose Dipole-lets, a dipole-adapted multiscale decomposition designed to identify phase components associated with streaking artifacts before and during susceptibility inversion. Dipole- lets combine an undecimated radial decomposition with an angular partition based on proximity to the magic cone. The resulting coefficients emphasize phase components near the magic cone, where nondipolar contributions may become relatively prominent. This information was incorporated into QSM reconstruction through a data- driven fidelity weight and a Dipole-let-based regularizer that models an additional nondipolar phase component. The methods were evaluated using modified simulated QSM data and an in vivo dataset. The data-driven weighting reduced streaking artifacts while preserving anatomical detail and improved quantitative reconstruction metrics compared with magnitude-weighted TV. The Dipole-let regularizer achieved streaking suppression comparable to L1-QSM and lower reconstruction error in the modified simulated dataset. In vivo experiments further demonstrated the applicability of the proposed approach to data with strong susceptibility-induced phase perturbations. Dipole-lets provide a multiscale representation for characterizing phase components associated with streaking artifacts and incorporating this information into QSM reconstruction, reducing streaking while preserving relevant susceptibility structures, and providing a flexible basis for future QSM reconstruction methods.

eess.IV↗
arXiv2026

MTMed3D: A Multi-Task Transformer-Based Model for 3D Medical Imaging

In the field of medical imaging, AI-assisted techniques such as object detection, segmentation, and classification are widely employed to alleviate the workload of physicians and doctors. However, single-task models are predominantly used, overlooking the shared information across tasks. This oversight leads to inefficiencies in real-life applications. In this work, we propose MTMed3D, a novel end-to-end Multi-task Transformer-based model to address the limitations of single-task models by jointly performing 3D detection, segmentation, and classification in medical imaging. Our model uses a Transformer as the shared encoder to generate multi-scale features, followed by CNN-based task-specific decoders. The proposed framework was evaluated on the BraTS 2018 and 2019 datasets, achieving promising results across all three tasks, especially in detection, where our method achieves better results than prior works. Additionally, we compare our multi-task model with equivalent single-task variants trained separately. Our multi-task model significantly reduces computational costs and achieves faster inference speed while maintaining comparable performance to the single-task models, highlighting its efficiency advantage. To the best of our knowledge, this is the first work to leverage Transformers for multi-task learning that simultaneously covers detection, segmentation, and classification tasks in 3D medical imaging, presenting its potential to enhance diagnostic processes. The code is available at https://github.com/fanlimua/MTMed3D.git.

eess.IV↗
arXiv2026

VideoPulse: Neonatal heart rate and peripheral capillary oxygen saturation (SpO2) estimation from contact free video

Remote photoplethysmography (rPPG) enables contact free monitoring of vital signs and is especially valuable for neonates, since conventional methods often require sustained skin contact with adhesive probes that can irritate fragile skin and increase infection control burden. We present VideoPulse, a neonatal dataset and an end to end pipeline that estimates neonatal heart rate and peripheral capillary oxygen saturation (SpO2) from facial video. VideoPulse contains 157 recordings totaling 2.6 hours from 52 neonates with diverse face orientations. Our pipeline performs face alignment and artifact aware supervision using denoised pulse oximeter signals, then applies 3D CNN backbones for heart rate and SpO2 regression with label distribution smoothing and weighted regression for SpO2. Predictions are produced in 2 second windows. On the NBHR neonatal dataset, we obtain heart rate MAE 2.97 bpm using 2 second windows (2.80 bpm at 6 second windows) and SpO2 MAE 1.69 percent. Under cross dataset evaluation, the NBHR trained heart rate model attains 5.34 bpm MAE on VideoPulse, and fine tuning an NBHR pretrained SpO2 model on VideoPulse yields MAE 1.68 percent. These results indicate that short unaligned neonatal video segments can support accurate heart rate and SpO2 estimation, enabling low cost non invasive monitoring in neonatal intensive care.

eess.IV↗