Transformer-based self-supervised models have achieved remarkable success in speech processing, but their large size and high inference cost present significant challenges for real-world deployment. While numerous compression techniques have been proposed, inconsistent evaluation metrics make it difficult to compare their practical effectiveness. In this work, we conduct a comprehensive study of four common compression methods, including weight pruning, head pruning, low-rank approximation, and knowledge distillation on self-supervised speech Transformers. We evaluate each method under three key metrics: parameter count, multiply-accumulate operations, and real-time factor. Results show that each method offers distinct advantages. In addition, we contextualize recent compression techniques, comparing DistilHuBERT, FitHuBERT, LightHuBERT, ARMHuBERT, and STaRHuBERT under the same framework, offering practical guidance on compression for deployment.

Occlusion remains a major challenge in online Multi-Object Tracking (MOT), where existing multi-stage association methods often rely on detection confidence scores despite their weak correlation with occlusion, leading to frequent errors. We propose DARUMA, a depth-aware MOT framework that prioritizes non-occluded objects using occlusion-aware association by re-prioritizing the matching order and refines association with a depth-weighted cost metric for improved robustness in occluded and depth-varying environments. Additionally, we introduce Generic Observation-Centric Momentum (GOCM), which integrates depth-aware velocity estimation and confidence-weighted historical observations to enhance motion modeling. Our method can integrates into existing MOT frameworks, improving association robustness without additional supervision.Extensive evaluations on DanceTrack demonstrate that DARUMA achieves state-of-the-art performance, particularly in complex, occlusion-heavy scenarios.

Token merging has emerged as an effective strategy to accelerate Vision Transformers (ViT) by reducing computational costs. However, existing methods primarily rely on the visual token’s feature similarity for token merging, overlooking the potential of integrating spatial information, which can serve as a reliable criterion for token merging in the early layers of ViT, where the visual tokens only possess weak visual information. In this paper, we propose ToSA, a novel token merging method that combines both semantic and spatial awareness to guide the token merging process. ToSA leverages the depth image as input to generate pseudo spatial tokens, which serve as auxiliary spatial information for the visual token merging process. With the introduced spatial awareness, ToSA achieves a more informed merging strategy that better preserves critical scene structure. Experimental results demonstrate that ToSA outperforms previous token merging methods across multiple benchmarks on visual and embodied question answering while largely reducing the runtime of the ViT, making it an efficient solution for ViT acceleration. The code will be available at: https://github.com/hsiangwei0903/ToSA.

Shih-Cheng Huang, Pin-Zu Li, Yu-chi Hsu, Kuang-Ming Chen, Yu Tung Lin, Shih-Kai Hsiao, Richard Tsai, Hung-yi Lee. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

This paper presents NFC+, a new approach to ubiquitous mobile media sharing. Data can be instantaneously shared in a fast and secure way between users without the presence of Internet. It uses NFC, which has fast setup time but low bandwidth, to exchange both network settings and the media URI, then programmatically configures wireless interfaces to securely connect and transfer the media. Specifically, we propose NFC+Hotspot, which combines NFC with secure Wi-Fi Hotspot mode. It is secure, easy to use, and significantly faster than existing sharing methods. We evaluated various sharing methods, in 2 categories, for both photo and video: 1. Direct Transfer (e.g. Bluetooth, Wi-Fi, Hotspot), and 2. NFC+ Direct Transfer. Experimental results show that our NFC+Hotspot approach has the fastest photo and video sharing performance, with improvement ranging from 1.8x-3.7x for 1MB photo and 0.7x -- 6.9x for 1 minute HD video compared to (non-NFC+) Direct Transfer methods.