USC SAIL's repositories
fed-multimodal
[KDD 2023] FedMultimodal
mica-MovieCLIP
This repository contains the codebase for MovieCLIP: Visual Scene Recognition in Movies
mica-subtitle-aligned-movie-sounds
A dataset for Audio-Visual Sound Event Detection in Movies
fed-ser-semi
Semi-supervised Learning for Speech Emotion Recognition On Federated Learning using Multiview Pseudo-Labeling
mica-screenplay-parser
Movie Screenplay Parser
tiles-dataset-release
Code accompanying the TILES dataset paper and data release. Data at https://tiles-data.isi.edu.
egocentric-fg-speech-detection
Egocentric Foreground Speech Detection
ggs_driving
Segmentation Algorithms for Physiological Time Series
mica-character-attribute-extraction
Character tropes, Forensic Interviews, and Character Attributes
mica-character-attribution
Learning descriptions of literary and fictional characters
mica-character-coref
Coreference in Movie Scripts
mica-context-emotion-recognition
Repository for context based emotion recognition
SCMIA-unsupervised-ASD
Repo for SCMIA and GSCMIA
tiles-audio-arousal
TILES speaking pattern (speaking frequency, duration, arousal rating) using audio data
tiles-time-series
The repo contains time-series implement for SAIL-TILES projects: preprocessing, segmentation, filtering, clustering, imputation, visualization...
CLAP
Contrastive Language-Audio Pretraining
HICA-active-speaker-detection
Repository for the HICA implementation associated with TMM submission
llama
Inference code for LLaMA models
mica-muse-india-tv
Data and code for analysis of the India TV Show Study
mica-muse-loreal
Codebase for analyzing RP metrics for MUSE LoREAL study
mica-scriptsonscreen-scripts
Contains code to scrape scriptsonscreen scripts website and scrapped data
tiles-2019-dataset
Code for the release of the TILES-2019 dataset