Vladimir Iashin

video_features. Extract video features from raw videos using multiple GPUs. We support RAFT flow frames as well as S3D, I3D, R(2+1)D, VGGish, CLIP, and TIMM models.

SpecVQGAN. Source code for "Taming Visually Guided Sound Generation" (Oral at the BMVC 2021)

BMT. Source code for "Bi-modal Transformer for Dense Video Captioning" (BMVC 2020)

MDVC. PyTorch implementation of Multi-modal Dense Video Captioning (CVPR 2020 Workshops)

Synchformer. Source code for "Synchformer: Efficient Synchronization from Sparse Cues" (ICASSP 2024)

SparseSync. Source code for "Sparse in Space and Time: Audio-visual Synchronisation with Trainable Selectors." (Spotlight at the BMVC 2022)

CS231n. PyTorch/Tensorflow solutions for Stanford's CS231n: "CNNs for Visual Recognition"

WebsiteYOLO. The back-end for the YOLOv3 object detector running as a webapp

VoxCeleb. An attempt to replicate the results of [1706.08612] VoxCeleb: a large-scale speaker identification dataset

CORSMAL. 🏆 🏆 Top-1 Submission to CORSMAL Challenge 2020 (at ICPR). The winning solution for the CORSMAL Challenge (on Intelligent Sensing Summer School 2020)

ChimpUFE. Universal Chimpanzee Face Embedding

FoursquareAPI. A simple example of using Foursquare API to get the data about a venue (Tips, UserId, and other info) and the data about a user (homecity, gender, number of friends, lists, checkins, photos, and tips) on Python

JumpMethod. Selecting a Proper Number of Clusters Using Jumps Method

v-iashin.github.io. Personal webpage

CrossEntropyTSP. An implementation of an approximation of the solution to Traveling Salesman Problem using cross entropy approach on Python 3

LearnablePINs. An attempt to replicate the results of [1805.00833] Learnable PINs: Cross-Modal Embeddings for Person Identity

TuniSurvivalKit. The ultimate survival kit for Ph.D. students at Tampere University

CopulaDensityEstimator. Recursive non-parametric estimation of the copula density

EM. EM-algorithm for two 1-dimentional Gaussians on vanilla-Python

SamplingChessboardWithQueens. Simulation of an N by M chessboard with K queens such that no queen defeats another using Simulated Annealing