2026-09-18T17:33:49Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1570062024-09-25T03:02:47Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023
Han, Song
Stiles, Nicole
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2024-09-24T18:26:22Z
2024-09-24T18:26:22Z
2024-05
2024-07-11T14:37:30.978Z
https://hdl.handle.net/1721.1/157006
https://orcid.org/0009-0001-2832-3552
The Segment-Anything Model (SAM) is a vision foundation model facilitating promptable and zero-shot image segmentation. SAM-based models have a wide range of applications including autonomous driving, medical image segmentation, VR, and data annotation. However, SAM models are highly computationally intensive and lack a flexible prompting mechanism. On an NVIDIA A100 GPU, SAM runs at 11 frames/second, missing the mark for real-time performance and preventing the usage of SAM on edge devices. To tackle both the latency constraint and the prompt flexibility constraint, we introduce GazeSAM, a new real-time gaze-prompted image segmentation model. GazeSAM uses face and gaze detection to determine the direction of a user's gaze, object detection to find candidate objects of interest, depth estimation to perform background detection, and image segmentation to generate masks. The final output is a mask segmenting the object at the focus of the user's gaze. By performing algorithmic optimizations, employing inference engines, and applying FP16 and INT8 quantization, we achieve a 24x speedup relative to the baseline FP32 PyTorch implementation. GazeSAM runs at a speed of over 30 FPS, enabling real-time performance on an RTX 4070 GPU.
M.Eng.
Massachusetts Institute of Technology
Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Copyright retained by author(s)
https://creativecommons.org/licenses/by-nc-nd/4.0/
Efficient Segment Anything on the Edge
Thesis
application/pdf
Master
Master of Engineering in Electrical Engineering and Computer Science
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="ee0f330f-35ed-48a7-99aa-0a387d58ed43">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Title>Efficient Segment Anything on the Edge</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2024-05</PublicationDate>
<Authors>
<Author>
<DisplayName>Stiles, Nicole</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>https://creativecommons.org/licenses/by-nc-nd/4.0/</License>
<Abstract>The Segment-Anything Model (SAM) is a vision foundation model facilitating promptable and zero-shot image segmentation. SAM-based models have a wide range of applications including autonomous driving, medical image segmentation, VR, and data annotation. However, SAM models are highly computationally intensive and lack a flexible prompting mechanism. On an NVIDIA A100 GPU, SAM runs at 11 frames/second, missing the mark for real-time performance and preventing the usage of SAM on edge devices. To tackle both the latency constraint and the prompt flexibility constraint, we introduce GazeSAM, a new real-time gaze-prompted image segmentation model. GazeSAM uses face and gaze detection to determine the direction of a user's gaze, object detection to find candidate objects of interest, depth estimation to perform background detection, and image segmentation to generate masks. The final output is a mask segmenting the object at the focus of the user's gaze. By performing algorithmic optimizations, employing inference engines, and applying FP16 and INT8 quantization, we achieve a 24x speedup relative to the baseline FP32 PyTorch implementation. GazeSAM runs at a speed of over 30 FPS, enabling real-time performance on an RTX 4070 GPU.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>