ABSTRACT Spoken Language Understanding (SLU), which aims to extract user semantics to execute downstream tasks, is a crucial component of task-oriented dialog systems. Existing SLU datasets generally lack sufficient diversity and complexity, and there is an absence of a unified benchmark for the latest Large Language Models (LLMs) and Large Audio Language Models (LALMs). This work introduces MAC-SLU, a novel Multi-Intent Automotive Cabin Spoken Language Understanding Dataset, which increases the difficulty of the SLU task by incorporating authentic and complex multi-intent data. Based on MAC-SLU, we conducted a comprehensive benchmark of leading open-source LLMs and LALMs, covering methods like in-context learning, supervised fine-tuning (SFT), and end-to-end (E2E) and pipeline paradigms. Our experiments show that while LLMs and LALMs have the potential to complete SLU tasks through in-context learning, their performance still lags significantly behind SFT. Meanwhile, E2E LALMs demonstrate performance comparable to pipeline approaches and effectively avoid error propagation from speech recognition.

The Yangwang U9 has become a sensation in the automotive world, not just for its impressive specifications but also for its remarkable capabilities. This electric hypercar from BYD's premium Yangwang brand boasts a staggering 1,287 horsepower and has recently been showcased performing extraordinary feats of autonomous driving. Enter the giveaway by signing up for my free newsletter. Launched with a price tag of approximately 236,000, the Yangwang U9 is designed to redefine performance in the electric vehicle sector. Equipped with four independent electric motors, it delivers an astonishing 1,287 hp and 1,680 Nm of torque, allowing it to accelerate from 0 to 60 mph in just 2.36 seconds.

Flying a drone is not easy as it seems. A lot of practice, patience, experience and concentration are all what required to fly a drone safely. EPFL has developed a system for drone control, taking away the sticks and replacing them with intuitive and comfortable movements of your entire body. Basically, it's an upper-body soft exoskeleton called FlyJacket. Developed by EPFL's Laboratory of Intelligent Systems, led by Professor Dario Floreano, FlyJacket is designed to be portable and affordable.