From 3e9be4a77b272749b0d692b1898b0dd81d520efd Mon Sep 17 00:00:00 2001 From: Dillon Laird Date: Wed, 11 Sep 2024 15:32:56 -0700 Subject: [PATCH] fix prompt for florence2 sam2 video tracking --- vision_agent/tools/tools.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/vision_agent/tools/tools.py b/vision_agent/tools/tools.py index 91e918d0..0e58049a 100644 --- a/vision_agent/tools/tools.py +++ b/vision_agent/tools/tools.py @@ -467,7 +467,8 @@ def florence2_sam2_video_tracking( entities in a video given a text prompt such as category names or referring expressions. You can optionally separate the categories in the text with commas. It only tracks entities present in the first frame and only returns segmentation - masks. It is useful for tracking and counting without duplicating counts. + masks. It is useful for tracking and counting without duplicating counts if they + appear in the first frame, always outputs scores of 1.0. Parameters: prompt (str): The prompt to ground to the video.