Result Details
Task: Deepfake Image/Video Detection
                Analysis Date: 2025-08-13 07:10:33.985405
                
                    
                        
                            Information:
Resolution: 848 x 480 | Codec: h264 | File Size: 1337.13KB | Bitrate: 219 kbps | Duration: 40.04 seconds | FPS: 25.0 |
                            
                        
                                
                                
                                
                                    
                                
                                
                                
                    
                
            Resolution: 848 x 480 | Codec: h264 | File Size: 1337.13KB | Bitrate: 219 kbps | Duration: 40.04 seconds | FPS: 25.0 |
                        Analytic Name: DSP-FWA
                        
                    
                                
                    
                            Description: Deep learning based method that can effectively distinguish AI-generated fake videos, this method target the artifacts in affine face warping as the distinctive feature to distinguish real and fake images
                            
                    
                                Reference:
                                
                                    https://arxiv.org/abs/1811.00656
                                
                            
                            
                                Code Link:
                                
                                    https://github.com/yuezunli/DSP-FWA
                                
                            
                        
                                
                                    🧪 Fake Probability: 71.54%
                                
                                
                                
                                
                                    ▲
                                    
                                
                                    71.54%
                                
                                
                                AI-generated Likelihood
                                
                        
                        
                        
                            Scope: Human face images or videos
                        
                    