arXiv subjects

Yury Kartynnik

Publications and source records attributed to Yury Kartynnik.

6 recordsLinked to original sources

arXiv2024

On Minimum Maximal Distance-k Matchings

We study the computational complexity of several problems connected with finding a maximal distance-$k$ matching of minimum cardinality or minimum weight in a given graph. We introduce the class of $k$-equimatchable graphs which is an edge analogue of $k$-equipackable graphs. We prove that the recognition of $k$-equimatchable graphs is co-NP-complete for any fixed $k \ge 2$. We provide a simple characterization for the class of strongly chordal graphs with equal $k$-packing and $k$-domination numbers. We also prove that for any fixed integer $\ell \ge 1$ the problem of finding a minimum weight maximal distance-$2\ell$ matching and the problem of finding a minimum weight $(2 \ell - 1)$-independent dominating set cannot be approximated in polynomial time in chordal graphs within a factor of $δ\ln |V(G)|$ unless $\mathrm{P} = \mathrm{NP}$, where $δ$ is a fixed constant (thereby improving the NP-hardness result of Chang for the independent domination case). Finally, we show the NP-hardness of the minimum maximal induced matching and independent dominating set problems in large-girth planar graphs. Note: This version (as compared to the journal submission) contains corrections to Section 4.

cs.DM↗

arXiv2024

StreamVC: Real-Time Low-Latency Voice Conversion

We present StreamVC, a streaming voice conversion solution that preserves the content and prosody of any source speech while matching the voice timbre from any target speech. Unlike previous approaches, StreamVC produces the resulting waveform at low latency from the input signal even on a mobile platform, making it applicable to real-time communication scenarios like calls and video conferencing, and addressing use cases such as voice anonymization in these scenarios. Our design leverages the architecture and training strategy of the SoundStream neural audio codec for lightweight high-quality speech synthesis. We demonstrate the feasibility of learning soft speech units causally, as well as the effectiveness of supplying whitened fundamental frequency information to improve pitch stability without leaking the source timbre information.

eess.AS↗

arXiv2020

Attention Mesh: High-fidelity Face Mesh Prediction in Real-time

We present Attention Mesh, a lightweight architecture for 3D face mesh prediction that uses attention to semantically meaningful regions. Our neural network is designed for real-time on-device inference and runs at over 50 FPS on a Pixel 2 phone. Our solution enables applications like AR makeup, eye tracking and AR puppeteering that rely on highly accurate landmarks for eye and lips regions. Our main contribution is a unified network architecture that achieves the same accuracy on facial landmarks as a multi-stage cascaded approach, while being 30 percent faster.

cs.CV↗

arXiv2019

Real-time Facial Surface Geometry from Monocular Video on Mobile GPUs

We present an end-to-end neural network-based model for inferring an approximate 3D mesh representation of a human face from single camera input for AR applications. The relatively dense mesh model of 468 vertices is well-suited for face-based AR effects. The proposed model demonstrates super-realtime inference speed on mobile GPUs (100-1000+ FPS, depending on the device and model variant) and a high prediction quality that is comparable to the variance in manual annotations of the same image.

cs.CV↗

arXiv2019

Real-time Hair Segmentation and Recoloring on Mobile GPUs

We present a novel approach for neural network-based hair segmentation from a single camera input specifically designed for real-time, mobile application. Our relatively small neural network produces a high-quality hair segmentation mask that is well suited for AR effects, e.g. virtual hair recoloring. The proposed model achieves real-time inference speed on mobile GPUs (30-100+ FPS, depending on the device) with high accuracy. We also propose a very realistic hair recoloring scheme. Our method has been deployed in major AR application and is used by millions of users.

cs.CV↗

arXiv2019

BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs

We present BlazeFace, a lightweight and well-performing face detector tailored for mobile GPU inference. It runs at a speed of 200-1000+ FPS on flagship devices. This super-realtime performance enables it to be applied to any augmented reality pipeline that requires an accurate facial region of interest as an input for task-specific models, such as 2D/3D facial keypoint or geometry estimation, facial features or expression classification, and face region segmentation. Our contributions include a lightweight feature extraction network inspired by, but distinct from MobileNetV1/V2, a GPU-friendly anchor scheme modified from Single Shot MultiBox Detector (SSD), and an improved tie resolution strategy alternative to non-maximum suppression.

cs.CV↗