Do You Really Mean That? Content Driven Audio-Visual Deepfake Dataset and Multimodal Method for Temporal Forgery Localization

13 Apr 2022  ·  Zhixi Cai, Kalin Stefanov, Abhinav Dhall, Munawar Hayat ·

Due to its high societal impact, deepfake detection is getting active attention in the computer vision community. Most deepfake detection methods rely on identity, facial attributes, and adversarial perturbation-based spatio-temporal modifications at the whole video or random locations while keeping the meaning of the content intact. However, a sophisticated deepfake may contain only a small segment of video/audio manipulation, through which the meaning of the content can be, for example, completely inverted from a sentiment perspective. We introduce a content-driven audio-visual deepfake dataset, termed Localized Audio Visual DeepFake (LAV-DF), explicitly designed for the task of learning temporal forgery localization. Specifically, the content-driven audio-visual manipulations are performed strategically to change the sentiment polarity of the whole video. Our baseline method for benchmarking the proposed dataset is a 3DCNN model, termed as Boundary Aware Temporal Forgery Detection (BA-TFD), which is guided via contrastive, boundary matching, and frame classification loss functions. Our extensive quantitative and qualitative analysis demonstrates the proposed method's strong performance for temporal forgery localization and deepfake detection tasks.

PDF Abstract


Introduced in the Paper:


Used in the Paper:


Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
DeepFake Detection LAV-DF BA-TFD AUC 0.990 # 1
Temporal Forgery Localization LAV-DF BA-TFD AR@100 66.9 # 3
AR@50 64.08 # 3
AR@20 60.77 # 3
AR@10 58.42 # 3
AP@0.5 76.9 # 3
AP@0.75 38.5 # 3
AP@0.95 0.25 # 3