vision-agent 0.2.126__py3-none-any.whl → 0.2.127__py3-none-any.whl

vision_agent/tools/tools.py

@@ -1518,7 +1518,9 @@ def save_video(
         raise ValueError(f"fps must be greater than 0 got {fps}")
 
     if output_video_path is None:
-        output_video_path = tempfile.NamedTemporaryFile(delete=False).name
+        output_video_path = tempfile.NamedTemporaryFile(
+            delete=False, suffix=".mp4"
+        ).name
 
     output_video_path = video_writer(frames, fps, output_video_path)
     _save_video_to_result(output_video_path)

vision_agent/utils/video.py

@@ -5,6 +5,7 @@ from functools import lru_cache
 from typing import List, Optional, Tuple
 
 import cv2
+import av  # type: ignore
 import numpy as np
 from decord import VideoReader  # type: ignore
 
@@ -43,18 +44,36 @@ def play_video(video_base64: str) -> None:
         cv2.destroyAllWindows()
 
 
+def _resize_frame(frame: np.ndarray) -> np.ndarray:
+    height, width = frame.shape[:2]
+    new_width = width - (width % 2)
+    new_height = height - (height % 2)
+    return cv2.resize(frame, (new_width, new_height))
+
+
 def video_writer(
     frames: List[np.ndarray], fps: float = 1.0, filename: Optional[str] = None
 ) -> str:
     if filename is None:
         filename = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4").name
-
-    fourcc = cv2.VideoWriter_fourcc(*"mp4v")  # type: ignore
+    container = av.open(filename, mode="w")
+    stream = container.add_stream("h264", rate=fps)
     height, width = frames[0].shape[:2]
-    writer = cv2.VideoWriter(filename, fourcc, fps, (width, height))
+    stream.height = height - (height % 2)
+    stream.width = width - (width % 2)
+    stream.pix_fmt = "yuv420p"
     for frame in frames:
-        writer.write(cv2.cvtColor(frame, cv2.COLOR_RGB2BGR))
-    writer.release()
+        # Remove the alpha channel (convert RGBA to RGB)
+        frame_rgb = frame[:, :, :3]
+        # Resize the frame to make dimensions divisible by 2
+        frame_rgb = _resize_frame(frame_rgb)
+        av_frame = av.VideoFrame.from_ndarray(frame_rgb, format="rgb24")
+        for packet in stream.encode(av_frame):
+            container.mux(packet)
+
+    for packet in stream.encode():
+        container.mux(packet)
+    container.close()
     return filename
 
 

{vision_agent-0.2.126.dist-info → vision_agent-0.2.127.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 0.2.126
+Version: 0.2.127
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai
@@ -10,6 +10,7 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Requires-Dist: anthropic (>=0.31.0,<0.32.0)
+Requires-Dist: av (>=11.0.0,<12.0.0)
 Requires-Dist: e2b (>=0.17.2a50,<0.18.0)
 Requires-Dist: e2b-code-interpreter (==0.0.11a37)
 Requires-Dist: eva-decord (>=0.6.1,<0.7.0)

{vision_agent-0.2.126.dist-info → vision_agent-0.2.127.dist-info}/RECORD

@@ -18,7 +18,7 @@ vision_agent/tools/__init__.py,sha256=T8Hi5aHf4J2QJDoPRvu5fxbiqMpAY-1Gi2EFIhJbf3
 vision_agent/tools/meta_tools.py,sha256=KeGiw2OtY8ARpGbtWjoNAoO1dwevt7LbCupaJX61MkE,18929
 vision_agent/tools/prompts.py,sha256=V1z4YJLXZuUl_iZ5rY0M5hHc_2tmMEUKr0WocXKGt4E,1430
 vision_agent/tools/tool_utils.py,sha256=62NVlojPMf9MuJ-3yJEcrB3mzmOxN2HrNQzzjVa-FZg,7527
-vision_agent/tools/tools.py,sha256=Y6BTLFoueLtjId2qG06UyZwCQA_TTA6uFxPkxzhRI50,65396
+vision_agent/tools/tools.py,sha256=sO0J-ts2CsJnf2UPcvxvmowE_G0X3f1iSChnS-cnPlk,65433
 vision_agent/tools/tools_types.py,sha256=rLpCUODPY0yI65SLOTJOxfHFfqWM3WjOq-AYX25Chjk,2356
 vision_agent/utils/__init__.py,sha256=7fMgbZiEwbNS0fBOS_hJI5PuEYBblw36zLi_UjUzvj4,244
 vision_agent/utils/exceptions.py,sha256=booSPSuoULF7OXRr_YbC4dtKt6gM_HyiFQHBuaW86C4,2052
@@ -26,8 +26,8 @@ vision_agent/utils/execute.py,sha256=gc4R_0BKUrZyhiKvIxOpYuzQPYVWQEqxr3ANy1lJAw4
 vision_agent/utils/image_utils.py,sha256=zTTOJFOieMzwIquTFnW7T6ssx9o6XfoZ0Unqyk7GJrg,10746
 vision_agent/utils/sim.py,sha256=ebE9Cs00pVEDI1HMjAzUBk88tQQmc2U-yAzIDinnekU,5572
 vision_agent/utils/type_defs.py,sha256=BE12s3JNQy36QvauXHjwyeffVh5enfcvd4vTzSwvEZI,1384
-vision_agent/utils/video.py,sha256=oM3sdQVGGI3xwrCN2GKt9otzDb0SPW-JUo5SABxTVl4,3847
-vision_agent-0.2.126.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-vision_agent-0.2.126.dist-info/METADATA,sha256=4O_OuQh5yhJ8unzNtfU4E_0RNykXxkbdjkiGPAXi9Ek,12258
-vision_agent-0.2.126.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
-vision_agent-0.2.126.dist-info/RECORD,,
+vision_agent/utils/video.py,sha256=oDTCuTv1dFMYvwqis7y0frt9U2iDF9KGN1g21bOVjvE,4528
+vision_agent-0.2.127.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+vision_agent-0.2.127.dist-info/METADATA,sha256=r3fKbSB79F3MsBsOTV0z054Qno3DTpf3Pa-xwkdIgD0,12295
+vision_agent-0.2.127.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
+vision_agent-0.2.127.dist-info/RECORD,,