Video quality enhancement#10
Open
masterface77 wants to merge 26 commits intoRafaelGodoyEbert:mainfrom
Open
Conversation
… in/hold/out/hold/repeat effect
… model, Portuguese default
…ng-Head - Add Auto Illumination (CLAHE) to video_quality.py - Pipeline: Denoise Auto Illumination Color Grading Unsharp - Add Blur Background mode (blurred bg + sharp center for 9:16) - Calibrate YOLO center_y offset (50% 30% for talking-head framing) - Update unsharp mask to Visual Opus params (5:5:1.0) - Add Colab notebook (viralcutter-Colab-Quality.ipynb) - Update Kaggle notebook with new pipeline - Update batch files with Visual Opus branding
… - InsightFace, TwoFace, Padding/Blur fallbacks now use enhance_frame(high) + Lanczos
…use wider 42% crop with blur background composite - Reduces upscale from 1.78x to 1.34x for 16:9 sources - Added 4K and 1440p download options (CLI, WebUI, download_video) - WebUI defaults to 4K quality
|
✅ Issue #10 Fechado Implementação de retry com recovery de jobs completada. Commit: 2517d21
Documentação: Ver API_DOCUMENTATION.md seção "Reprocessar Job" e |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Olá! 👋
Este PR introduz melhorias significativas no sistema de tracking facial e na compatibilidade com ambientes de nuvem (Colab e Kaggle). O foco principal é a implementação do YOLO Smooth Tracking, que substitui cortes bruscos por um movimento de câmera suave e cinematográfico, além de otimizações de memória para GPUs T4.
🚀 Principais Alterações
1. Smooth Face Tracking (Cinematográfico)
Implementação de YOLOv8 + ByteTrack para detecção e persistência de IDs.
Adição de EMA (Exponential Moving Average) para suavização do movimento da câmera.
Controle no Gradio: Nova aba "Advanced Face Settings" com slider de suavidade (
Tracking Smoothness).0.02 (Ultra Suave) a 0.10 (Rápido).
2. Otimização para GPUs T4 (Colab/Kaggle)
Ajustes no WhisperX para evitar OOM (Out of Memory) em GPUs de 16GB:
compute_type="int8_float16"(reduz uso de VRAM em ~50%).batch_size=8.chunk_size=15.3. Novos Notebooks e Suporte a Kaggle
Notebook Colab Smooth Zoom: Configuração facilitada com upload automático para o Drive.
Notebook Kaggle: Suporte completo adicionado (aproveitando as 30h semanais de GPU grátis).
Inclui instruções para setup de Datasets (Client Secret, Cookies, Gemini API).
Upload via OAuth implementado.
📊 Comparativo de Ambientes Adicionados
Característica | Colab | Kaggle -- | -- | -- Zoom IA | Removido (nesta versão) | ✅ Disponível Persistência | Nenhuma | ✅ Via Datasets GPU Grátis | ~12h/dia (instável) | 30h/semana (estável)🧪 Como testar
Utilizar os notebooks incluídos (
ViralCutter-SmoothZoom.ipynbou versão Kaggle).No Gradio, selecionar o Face Model como
yolo.Ajustar o slider "Tracking Smoothness" para verificar a diferença na suavidade do vídeo gerado.