Video Token Sparsification for Efficient Multimodal LLMs in Autonomous Driving
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914951009927168 |
|---|---|
| author | Ma, Yunsheng Abdelraouf, Amr Gupta, Rohit Wang, Ziran Han, Kyungtae |
| author_facet | Ma, Yunsheng Abdelraouf, Amr Gupta, Rohit Wang, Ziran Han, Kyungtae |
| contents | Multimodal large language models (MLLMs) have demonstrated remarkable potential for enhancing scene understanding in autonomous driving systems through powerful logical reasoning capabilities. However, the deployment of these models faces significant challenges due to their substantial parameter sizes and computational demands, which often exceed the constraints of onboard computation. One major limitation arises from the large number of visual tokens required to capture fine-grained and long-context visual information, leading to increased latency and memory consumption. To address this issue, we propose Video Token Sparsification (VTS), a novel approach that leverages the inherent redundancy in consecutive video frames to significantly reduce the total number of visual tokens while preserving the most salient information. VTS employs a lightweight CNN-based proposal model to adaptively identify key frames and prune less informative tokens, effectively mitigating hallucinations and increasing inference throughput without compromising performance. We conduct comprehensive experiments on the DRAMA and LingoQA benchmarks, demonstrating the effectiveness of VTS in achieving up to a 33\% improvement in inference throughput and a 28\% reduction in memory usage compared to the baseline without compromising performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_11182 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Video Token Sparsification for Efficient Multimodal LLMs in Autonomous Driving Ma, Yunsheng Abdelraouf, Amr Gupta, Rohit Wang, Ziran Han, Kyungtae Computer Vision and Pattern Recognition Multimodal large language models (MLLMs) have demonstrated remarkable potential for enhancing scene understanding in autonomous driving systems through powerful logical reasoning capabilities. However, the deployment of these models faces significant challenges due to their substantial parameter sizes and computational demands, which often exceed the constraints of onboard computation. One major limitation arises from the large number of visual tokens required to capture fine-grained and long-context visual information, leading to increased latency and memory consumption. To address this issue, we propose Video Token Sparsification (VTS), a novel approach that leverages the inherent redundancy in consecutive video frames to significantly reduce the total number of visual tokens while preserving the most salient information. VTS employs a lightweight CNN-based proposal model to adaptively identify key frames and prune less informative tokens, effectively mitigating hallucinations and increasing inference throughput without compromising performance. We conduct comprehensive experiments on the DRAMA and LingoQA benchmarks, demonstrating the effectiveness of VTS in achieving up to a 33\% improvement in inference throughput and a 28\% reduction in memory usage compared to the baseline without compromising performance. |
| title | Video Token Sparsification for Efficient Multimodal LLMs in Autonomous Driving |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2409.11182 |