Update readme for hf

README.md

@@ -1,16 +1,22 @@
+---
+license: apache-2.0
+datasets:
+- laion/LAION-DISCO-12M
+language:
+- en
+- zh
+base_model:
+- stabilityai/stable-audio-open-1.0
+pipeline_tag: audio-to-audio
+tags:
+- music
+- vae
+---
 # εar-VAE: High Fidelity Music Reconstruction Model
+[[Demo Page](https://eps-acoustic-revolution-lab.github.io/EAR_VAE/)] - [[Codes](https://github.com/Eps-Acoustic-Revolution-Lab/EAR_VAE)]
 
 This repository contains the official inference code for εar-VAE, aa 44.1 kHz music signal reconstruction model that rethinks and optimizes VAE training for audio. It targets two common weaknesses in existing open-source VAEs—phase accuracy and stereophonic spatial representation—by aligning objectives with auditory perception and introducing phase-aware training. Experiments show substantial improvements across diverse metrics, with particular strength in high-frequency harmonics and spatial characteristics.
 
-<p align="center">
-<img src="./images/all_compares.jpg" width=90%>
-<img src="./images/table.png" width=90%>
-</p>
-
-<p align="center">
-<em>Upper: Ablation study across our training components.</em> <em>Down: Cross-model metric comparison on the evaluation dataset.</em>
-</p>
-
 Why εar-VAE:
 - 🎧 Perceptual alignment: A K-weighting perceptual filter is applied before loss computation to better match human hearing.
 - 🔁 Phase-aware objectives: Two novel phase losses
@@ -142,6 +148,4 @@ This project builds upon the work of several open-source projects. We would like
 -   **[Stability AI's Stable Audio Tools](https://github.com/Stability-AI/stable-audio-tools)**: For providing a foundational framework and tools for audio generation.
 -   **[Descript's Audio Codec](https://github.com/descriptinc/descript-audio-codec)**: For the weight-normed convolusional layers
 
-Their contributions have been invaluable to the development of εar-VAE.
-
-
+Their contributions have been invaluable to the development of εar-VAE.