Heralded as the ChatGPT for music, Suno AI is the latest iteration of generative artificial intelligence to flood social feeds, wowing users with its (ahem) lyrical prowess. Plug in the musical style you want, a genre and a prompt for lyrics and Suno can spit out a full song for you in a matter of seconds. The business has been around for two years, formulated by a group of machine learning experts in Cambridge who struck an interest in audio, according to a profile in Rolling Stone last month. From the outset, making silly songs is slightly addictive. The lyrics might seem shallow and soulless, but they're also often hilarious.

As AI-generated image (AIGI) methods become more powerful and accessible, it has become a critical task to determine if an image is real or AI-generated. Because AIGI lack the signatures of photographs and have their own unique patterns, new models are needed to determine if an image is AI-generated. In this paper, we investigate the ability of the Contrastive Language-Image Pre-training (CLIP) architecture, pre-trained on massive internet-scale data sets, to perform this differentiation. We fine-tune CLIP on real images and AIGI from several generative models, enabling CLIP to determine if an image is AI-generated and, if so, determine what generation method was used to create it. We show that the fine-tuned CLIP architecture is able to differentiate AIGI as well or better than models whose architecture is specifically designed to detect AIGI. Our method will significantly increase access to AIGI-detecting tools and reduce the negative effects of AIGI on society, as our CLIP fine-tuning procedures require no architecture changes from publicly available model repositories and consume significantly less GPU resources than other AIGI detection models.

Recent advancement in large language models (LLMs) has offered a strong potential for natural language systems to process informal language. A representative form of informal language is slang, used commonly in daily conversations and online social media. To date, slang has not been comprehensively evaluated in LLMs due partly to the absence of a carefully designed and publicly accessible benchmark. Using movie subtitles, we construct a dataset that supports evaluation on a diverse set of tasks pertaining to automatic processing of slang. For both evaluation and finetuning, we show the effectiveness of our dataset on two core applications: 1) slang detection, and 2) identification of regional and historical sources of slang from natural sentences. We also show how our dataset can be used to probe the output distributions of LLMs for interpretive insights. We find that while LLMs such as GPT-4 achieve good performance in a zero-shot setting, smaller BERT-like models finetuned on our dataset achieve comparable performance. Furthermore, we show that our dataset enables finetuning of LLMs such as GPT-3.5 that achieve substantially better performance than strong zero-shot baselines. Our work offers a comprehensive evaluation and a high-quality benchmark on English slang based on the OpenSubtitles corpus, serving both as a publicly accessible resource and a platform for applying tools for informal language processing.

Radio Frequency Fingerprint Identification (RFFI), which exploits non-ideal hardware-induced unique distortion resident in the transmit signals to identify an emitter, is emerging as a means to enhance the security of communication systems. Recently, machine learning has achieved great success in developing state-of-the-art RFFI models. However, few works consider cross-receiver RFFI problems, where the RFFI model is trained and deployed on different receivers. Due to altered receiver characteristics, direct deployment of RFFI model on a new receiver leads to significant performance degradation. To address this issue, we formulate the cross-receiver RFFI as a model adaptation problem, which adapts the trained model to unlabeled signals from a new receiver. We first develop a theoretical generalization error bound for the adaptation model. Motivated by the bound, we propose a novel method to solve the cross-receiver RFFI problem, which includes domain alignment and adaptive pseudo-labeling. The former aims at finding a feature space where both domains exhibit similar distributions, effectively reducing the domain discrepancy. Meanwhile, the latter employs a dynamic pseudo-labeling scheme to implicitly transfer the label information from the labeled receiver to the new receiver. Experimental results indicate that the proposed method can effectively mitigate the receiver impact and improve the cross-receiver RFFI performance.

My editors yell at me for over-using it. So it's natural that I would latch onto Udio.com, a surprisingly capable AI music generator for everything from death metal to female folk rock. The thing is, I'm not sure how long it will be as great as it is right now. Udio works like an AI art generator. You can either specify a prompt and let Udio do all the heavy lifting, from music to lyrics, or get as detailed as you want.

For more than a decade, blockbuster comic book adaptations reliably clobbered all competition at the box office. Disney and HBO Max built their streaming strategies around intellectual property from Marvel and DC Comics. The studios turned this pulpy source material into a profusion of interconnected films and series that consistently drove ticket sales and subscriptions--until they didn't. Lately, serious superhero fatigue seems to have set in. Comic book movies regularly tank these days, and not just the ones based on second-string characters like Blue Beetle and Madame Web.

I've found myself at a loss for words when trying to explain the Humane AI Pin to my friends. The best description so far is that it's a combination of a wearable Siri button with a camera and built-in projector that beams onto your palm. But each time I start explaining that, I get so caught up in pointing out its problems that I never really get to fully detail what the AI Pin can do. Or is meant to do, anyway. Yet, words are crucial to the Humane AI experience. Your primary mode of interacting with the pin is through voice, accompanied by touch and gestures. Without speaking, your options are severely limited. The company describes the device as your "second brain," but the combination of holding out my hand to see the projected screen, waving it around to navigate the interface and tapping my chest and waiting for an answer all just made me look really stupid. When I remember that I was actually eager to spend 700 of my own money to get a Humane AI Pin, not to mention shell out the required 24 a month for the AI and the company's 4G service riding on T-Mobile's network, I feel even sillier. In the company's own words, the Humane AI Pin is the "first wearable device and software platform built to harness the full power of artificial intelligence." There are basically two parts to the device: the Pin and its magnetic attachment.

We present a corpus of 100 documents, named OBSINFOX, selected from 17 sources of French press considered unreliable by expert agencies, annotated using 11 labels by 8 annotators. By collecting more labels than usual, by more annotators than is typically done, we can identify features that humans consider as characteristic of fake news, and compare them to the predictions of automated classifiers. We present a topic and genre analysis using GATE Cloud, indicative of the prevalence of satire-like text in the corpus. We then use the subjectivity analyzer VAGO, and a neural version of it, to clarify the link between ascriptions of the label Subjective and ascriptions of the label Fake News.

The recent advancement in deep learning (DL) for automatic modulation classification (AMC) of wireless signals has encouraged numerous possible applications on resource-constrained edge devices. However, developing optimized DL models suitable for edge applications of wireless communications is yet to be studied in depth. In this work, we perform a thorough investigation of optimized convolutional neural networks (CNNs) developed for AMC using the three most commonly used model optimization techniques: a) pruning, b) quantization, and c) knowledge distillation. Furthermore, we have proposed optimized models with the combinations of these techniques to fuse the complementary optimization benefits. The performances of all the proposed methods are evaluated in terms of sparsity, storage compression for network parameters, and the effect on classification accuracy with a reduction in parameters. The experimental results show that the proposed individual and combined optimization techniques are highly effective for developing models with significantly less complexity while maintaining or even improving classification performance compared to the benchmark CNNs.

In this position paper, we discuss the potential for leveraging LLMs as interactive research tools to facilitate collaboration between human coders and AI to effectively annotate online risk data at scale. Collaborative human-AI labeling is a promising approach to annotating large-scale and complex data for various tasks. Yet, tools and methods to support effective human-AI collaboration for data annotation are under-studied. This gap is pertinent because co-labeling tasks need to support a two-way interactive discussion that can add nuance and context, particularly in the context of online risk, which is highly subjective and contextualized. Therefore, we provide some of the early benefits and challenges of using LLMs-based tools for risk annotation and suggest future directions for the HCI research community to leverage LLMs as research tools to facilitate human-AI collaboration in contextualized online data annotation. Our research interests align very well with the purposes of the LLMs as Research Tools workshop to identify ongoing applications and challenges of using LLMs to work with data in HCI research. We anticipate learning valuable insights from organizers and participants into how LLMs can help reshape the HCI community's methods for working with data.