Germany

Marcella Astrid

Advanced
@aseuteurideu

AI consultant at Helmholtz AI, Helmholtz Zentrum München

STEAL. Official implementation of "Synthetic Temporal Anomaly Guided End-to-End Video Anomaly Detection" (ICCV Workshops 2021: RSL-CV).

29

ADNet_pytorch. ADNet implementation on pytorch

24

LearningNotToReconstructAnomalies. Official implementation of "Learning Not to Reconstruct" (BMVC 2021)

19

FGI. Detecting Audio-Visual Deepfakes with Fine-Grained Inconsistencies

7

TADA. Targeted Augmented Data for Audio Deepfake Detection

4

LTI. Audio-visual Deepfake Detection With Local Temporal Inconsistencies (ICASSP 2025)

3

CP-TPM-CNN. The CNN weight from "CP-decomposition with Tensor Power Method for Convolutional Neural Networks compression" (http://ieeexplore.ieee.org/document/7881725/)

2

CP-TPM-CNNv2. Continuation of CP-TPM-CNN.. Based on my master thesis https://www.researchgate.net/publication/319623885_Deep_Compression_of_Convolutional_Neural_Networks_with_Low-Rank_Approximation

1

SADD. Code for Statistics-aware Audio-visual Deepfake Detector

1
9
Apply