rm cite

app.py

@@ -18,25 +18,25 @@ from PIL import Image
 
 
 TRANSLATE = {
-    "Symphony": "交响乐 Symphony",
-    "Opera": "戏曲 Opera",
-    "Solo": "独奏 Solo",
-    "Chamber": "室内乐 Chamber",
-    "Pop_vocal_ballad": "芭乐 Pop vocal ballad",
-    "Adult_contemporary": "成人时代 Adult contemporary",
-    "Teen_pop": "青少年流行 Teen pop",
-    "Contemporary_dance_pop": "当代流行舞曲 Contemporary dance pop",
-    "Dance_pop": "流行舞曲 Dance pop",
-    "Classic_indie_pop": "经典独立流行 Classic indie pop",
-    "Chamber_cabaret_and_art_pop": "室内卡巴莱与艺术流行乐 Chamber cabaret & art pop",
-    "Soul_or_r_and_b": "灵魂乐或节奏布鲁斯 Soul / R&B",
-    "Adult_alternative_rock": "成人另类摇滚 Adult alternative rock",
-    "Uplifting_anthemic_rock": "迷幻民族摇滚 Uplifting anthemic rock",
-    "Soft_rock": "慢摇滚 Soft rock",
-    "Acoustic_pop": "原声流行 Acoustic pop",
+    "Symphony": "Symphony",
+    "Opera": "Opera",
+    "Solo": "Solo",
+    "Chamber": "Chamber",
+    "Pop_vocal_ballad": "Pop vocal ballad",
+    "Adult_contemporary": "Adult contemporary",
+    "Teen_pop": "Teen pop",
+    "Contemporary_dance_pop": "Contemporary dance pop",
+    "Dance_pop": "Dance pop",
+    "Classic_indie_pop": "Classic indie pop",
+    "Chamber_cabaret_and_art_pop": "Chamber cabaret & art pop",
+    "Soul_or_r_and_b": "Soul / R&B",
+    "Adult_alternative_rock": "Adult alternative rock",
+    "Uplifting_anthemic_rock": "Uplifting anthemic rock",
+    "Soft_rock": "Soft rock",
+    "Acoustic_pop": "Acoustic pop",
 }
-
 CLASSES = list(TRANSLATE.keys())
+CACHE_DIR = "__pycache__"
 
 
 def most_common_element(input_list):
@@ -46,7 +46,7 @@ def most_common_element(input_list):
 
 
 def mp3_to_mel(audio_path: str, width=11.4):
-    os.makedirs("./flagged", exist_ok=True)
+    os.makedirs(CACHE_DIR, exist_ok=True)
     try:
         y, sr = librosa.load(audio_path)
         mel_spec = librosa.feature.melspectrogram(y=y, sr=sr)
@@ -61,7 +61,7 @@ def mp3_to_mel(audio_path: str, width=11.4):
             librosa.display.specshow(log_mel_spec[:, i : i + step])
             plt.axis("off")
             plt.savefig(
-                f"./flagged/mel_{round(dur, 2)}_{i}.jpg",
+                f"{CACHE_DIR}/mel_{round(dur, 2)}_{i}.jpg",
                 bbox_inches="tight",
                 pad_inches=0.0,
             )
@@ -72,7 +72,7 @@ def mp3_to_mel(audio_path: str, width=11.4):
 
 
 def mp3_to_cqt(audio_path: str, width=11.4):
-    os.makedirs("./flagged", exist_ok=True)
+    os.makedirs(CACHE_DIR, exist_ok=True)
     try:
         y, sr = librosa.load(audio_path)
         cqt_spec = librosa.cqt(y=y, sr=sr)
@@ -87,7 +87,7 @@ def mp3_to_cqt(audio_path: str, width=11.4):
             librosa.display.specshow(log_cqt_spec[:, i : i + step])
             plt.axis("off")
             plt.savefig(
-                f"./flagged/cqt_{round(dur, 2)}_{i}.jpg",
+                f"{CACHE_DIR}/cqt_{round(dur, 2)}_{i}.jpg",
                 bbox_inches="tight",
                 pad_inches=0.0,
             )
@@ -98,7 +98,7 @@ def mp3_to_cqt(audio_path: str, width=11.4):
 
 
 def mp3_to_chroma(audio_path: str, width=11.4):
-    os.makedirs("./flagged", exist_ok=True)
+    os.makedirs(CACHE_DIR, exist_ok=True)
     try:
         y, sr = librosa.load(audio_path)
         chroma_spec = librosa.feature.chroma_stft(y=y, sr=sr)
@@ -113,7 +113,7 @@ def mp3_to_chroma(audio_path: str, width=11.4):
             librosa.display.specshow(log_chroma_spec[:, i : i + step])
             plt.axis("off")
             plt.savefig(
-                f"./flagged/chroma_{round(dur, 2)}_{i}.jpg",
+                f"{CACHE_DIR}/chroma_{round(dur, 2)}_{i}.jpg",
                 bbox_inches="tight",
                 pad_inches=0.0,
             )
@@ -135,12 +135,12 @@ def embed_img(img_path, input_size=224):
     return transform(img).unsqueeze(0)
 
 
-def inference(mp3_path, log_name: str, folder_path="./flagged"):
+def inference(mp3_path, log_name: str, folder_path=CACHE_DIR):
     if os.path.exists(folder_path):
         shutil.rmtree(folder_path)
 
     if not mp3_path:
-        return None, "请输入音频 Please input an audio!"
+        return None, "Please input an audio!"
 
     network = EvalNet(log_name)
     spec = log_name.split("_")[-1]
@@ -186,35 +186,17 @@ if __name__ == "__main__":
         gr.Interface(
             fn=inference,
             inputs=[
-                gr.Audio(label="上传MP3音频 Upload MP3", type="filepath"),
-                gr.Dropdown(
-                    choices=models, label="选择模型 Select a model", value=models[6]
-                ),
+                gr.Audio(label="Upload MP3", type="filepath"),
+                gr.Dropdown(choices=models, label="Select a model", value=models[6]),
             ],
             outputs=[
-                gr.Textbox(label="音频文件名 Audio filename", show_copy_button=True),
-                gr.Textbox(label="流派识别 Genre recognition", show_copy_button=True),
+                gr.Textbox(label="Audio filename", show_copy_button=True),
+                gr.Textbox(label="Genre recognition", show_copy_button=True),
             ],
             examples=examples,
             cache_examples=False,
             allow_flagging="never",
-            title="建议录音时长保持在 15s 以内, 过长会影响识别效率<br>It is recommended to keep the duration of recording within 15s, too long will affect the recognition efficiency.",
-        )
-
-        gr.Markdown(
-            """
-# 引用 Cite
-```bibtex
-@dataset{zhaorui_liu_2021_5676893,
-  author       = {Monan Zhou, Shenyang Xu, Zhaorui Liu, Zhaowen Wang, Feng Yu, Wei Li and Baoqiang Han},
-  title        = {CCMusic: an Open and Diverse Database for Chinese and General Music Information Retrieval Research},
-  month        = {mar},
-  year         = {2024},
-  publisher    = {HuggingFace},
-  version      = {1.2},
-  url          = {https://huggingface.co/ccmusic-database}
-}
-```"""
+            title="It is recommended to keep the duration of recording within 15s, too long will affect the recognition efficiency.",
         )
 
     demo.launch()

requirements.txt

@@ -1,6 +1,6 @@
-librosa
 torch
+pillow
+librosa
 matplotlib
 torchvision
-pillow
-modelscope==1.15
+modelscope[framework]==1.18

utils.py

@@ -32,23 +32,17 @@ def get_modelist(model_dir=MODEL_DIR):
     try:
         entries = os.listdir(model_dir)
     except OSError as e:
-        print(f"无法访问 {model_dir}: {e}")
+        print(f"Cannot access {model_dir}: {e}")
         return
 
-    # 遍历所有条目
     output = []
     for entry in entries:
-        # 获取完整路径
         full_path = os.path.join(model_dir, entry)
-
-        # 跳过'.git'文件夹
         if entry == ".git" or entry == "examples":
-            print(f"跳过 .git / examples 文件夹: {full_path}")
+            print(f"Skip .git / examples dir: {full_path}")
             continue
 
-        # 检查条目是文件还是目录
         if os.path.isdir(full_path):
-            # 打印目录路径
             output.append(os.path.basename(full_path))
 
     return output
@@ -62,6 +56,6 @@ def download(url: str):
             for chunk in response.iter_content(chunk_size=8192):
                 f.write(chunk)
 
-        print(f"文件已下载到 {os.getcwd()}/{filename}")
+        print(f"The file has been downloaded to {os.getcwd()}/{filename}")
     else:
-        print(f"下载失败，状态码：{response.status_code}")
+        print(f"Failed to download, status code: {response.status_code}")