Ouranalysis include examination of the cross-attention maps of the learned prompts at individual time steps of the denoising process.

Recently, prompt learning has emerged as the state-of-the-art (SOT A) for fair text-to-image (T2I) generation.

In this paper, we introduce a speculative design methodology for studying the behavior of generative AI systems, framing design as a mode of inquiry. We propose bridging seemingly unrelated domains to generate intentional context voids, using these tasks as probes to elicit AI model behavior. We demonstrate this through a case study: probing the ChatGPT system (GPT-4 and DALL-E) to generate headshots from professional Curricula Vitae (CVs). In contrast to traditional ways, our approach assesses system behavior under conditions of radical uncertainty -- when forced to invent entire swaths of missing context -- revealing subtle stereotypes and value-laden assumptions. We qualitatively analyze how the system interprets identity and competence markers from CVs, translating them into visual portraits despite the missing context (i.e. physical descriptors). We show that within this context void, the AI system generates biased representations, potentially relying on stereotypical associations or blatant hallucinations.

Recently, prompt learning has emerged as the state-of-the-art (SOTA) for fair text-to-image (T2I) generation. Specifically, this approach leverages readily available reference images to learn inclusive prompts for each target Sensitive Attribute (tSA), allowing for fair image generation. In this work, we first reveal that this prompt learning-based approach results in degraded sample quality. Our analysis shows that the approach's training objective -- which aims to align the embedding differences of learned prompts and reference images -- could be sub-optimal, resulting in distortion of the learned prompts and degraded generated images. To further substantiate this claim, as our major contribution, we deep dive into the denoising subnetwork of the T2I model to track down the effect of these learned prompts by analyzing the cross-attention maps. In our analysis, we propose a novel prompt switching analysis: I2H and H2I. Furthermore, we propose new quantitative characterization of cross-attention maps. Our analysis reveals abnormalities in the early denoising steps, perpetuating improper global structure that results in degradation in the generated samples. Building on insights from our analysis, we propose two ideas: (i) Prompt Queuing and (ii) Attention Amplification to address the quality issue. Extensive experimental results on a wide range of tSAs show that our proposed method outperforms SOTA approach's image generation quality, while achieving competitive fairness. More resources at FairQueue Project site: https://sutd-visual-computing-group.github.io/FairQueue

Text-to-image generative models often present issues regarding fairness with respect to certain sensitive attributes, such as gender or skin tone. This study aims to reproduce the results presented in "ITI-Gen: Inclusive Text-to-Image Generation" by Zhang et al. (2023a), which introduces a model to improve inclusiveness in these kinds of models. We show that most of the claims made by the authors about ITI-Gen hold: it improves the diversity and quality of generated images, it is scalable to different domains, it has plug-and-play capabilities, and it is efficient from a computational point of view. However, ITI-Gen sometimes uses undesired attributes as proxy features and it is unable to disentangle some pairs of (correlated) attributes such as gender and baldness. In addition, when the number of considered attributes increases, the training time grows exponentially and ITI-Gen struggles to generate inclusive images for all elements in the joint distribution. To solve these issues, we propose using Hard Prompt Search with negative prompting, a method that does not require training and that handles negation better than vanilla Hard Prompt Search. Nonetheless, Hard Prompt Search (with or without negative prompting) cannot be used for continuous attributes that are hard to express in natural language, an area where ITI-Gen excels as it is guided by images during training. Finally, we propose combining ITI-Gen and Hard Prompt Search with negative prompting.

When they finally appeared, I squinted, bewildered. I laughed out loud alone in my office, then sent a very confused email to customer support. Every writer I know is talking about AI tools and whether they're ethical to use. But what's just as interesting to me is why we're so enamored with them even if they produce outrageous or below-average results. Why do we chase the shiny new thing even if it's not better, faster, or cheaper?

Text-to-image generative AI models such as Stable Diffusion are used daily by millions worldwide. However, many have raised concerns regarding how these models amplify racial and gender stereotypes. To study this phenomenon, we develop a classifier to predict the race, gender, and age group of any given face image, and show that it achieves state-of-the-art performance. Using this classifier, we quantify biases in Stable Diffusion across six races, two genders, five age groups, 32 professions, and eight attributes. We then propose novel debiasing solutions that outperform state-of-the-art alternatives. Additionally, we examine the degree to which Stable Diffusion depicts individuals of the same race as being similar to one another. This analysis reveals a high degree of stereotyping, e.g., depicting most middle eastern males as being dark-skinned, bearded, and wearing a traditional headdress. We address these limitations by proposing yet another novel solution that increases facial diversity across genders and racial groups. Our solutions are open-sourced and made publicly available.

Phantom Liberty is CD Projekt RED's masterpiece. Not only is Cyberpunk 2077 Phantom Liberty graphically easily three generations ahead of the entire industry and redefines how we experience video games with pathtracing, it's also written even more thrillingly and staged even more explosively. Anyone who doesn't enjoy this several times in different play styles has never loved video games. Cyberpunk 2077's Phantom Liberty expansion is a reminder of how incredibly explosive gaming has become – and the perfection with which CD Projekt RED manages to involve its actors. When Idris Elba is on a train out of Dogtown, joking with Songbird about how they really need to eat that one famous burrito of his together sometime, and there's such an eerie silence to the flirtation – the nervous looks of the head hacker because she's about to betray him – these are moments that feel like they'd belong in House of Cards or 24.

Text-to-image generative models often reflect the biases of the training data, leading to unequal representations of underrepresented groups. This study investigates inclusive text-to-image generative models that generate images based on human-written prompts and ensure the resulting images are uniformly distributed across attributes of interest. Unfortunately, directly expressing the desired attributes in the prompt often leads to sub-optimal results due to linguistic ambiguity or model misrepresentation. Hence, this paper proposes a drastically different approach that adheres to the maxim that "a picture is worth a thousand words". We show that, for some attributes, images can represent concepts more expressively than text. For instance, categories of skin tones are typically hard to specify by text but can be easily represented by example images. Building upon these insights, we propose a novel approach, ITI-GEN, that leverages readily available reference images for Inclusive Text-to-Image GENeration. The key idea is learning a set of prompt embeddings to generate images that can effectively represent all desired attribute categories. More importantly, ITI-GEN requires no model fine-tuning, making it computationally efficient to augment existing text-to-image models. Extensive experiments demonstrate that ITI-GEN largely improves over state-of-the-art models to generate inclusive images from a prompt. Project page: https://czhang0528.github.io/iti-gen.

Conservative women are more attractive than left-wing females, according to a European study of thousands of faces. Danish and Swedish researchers tested a deep-learning artificial intelligence, called a neural network, that can predict a person's political leanings the majority of the time, based solely on their headshot. It found that right wing women more were attractive, based on a publicly available scoring system. The group found no such link in men, but did determine that the left-leaning men showed more neutral, less happy faces, suggesting perhaps better skill at guarding their emotions. The true purpose of the researchers' study, however, was to show the alarming accuracy of off-the shelf AI, which can correctly guess a person's political views based on limited information, like a simple selfie, posted to social media everyday.