AI Struggles to Capture Disgust: A Look at Facial Expressions in Generated Images with Flux-dev
- 10 minutes read - 2056 words
Table of Contents
Facial expressions are a powerful tool for conveying emotions, and disgust is one of the most visually distinct. It’s characterized by a unique combination of facial contortions, including wrinkled noses, raised upper lips, and furrowed brows. These expressions are often accompanied by a sense of revulsion or aversion, making them particularly challenging for AI models to capture accurately. This blog post delves into the results of an experiment that tested the ability of a generative AI model to depict disgust in various scenarios, highlighting the model’s strengths and weaknesses in capturing this complex human emotion.
Created with: flux-dev
Intense Gaze in a Formal Setting: A Moment of Political Tension
A man in a suit, his expression serious and focused, stares intently at another figure in a room adorned with a chandelier and wood-paneled walls. The scene exudes a palpable sense of tension and political intrigue, leaving the viewer to wonder what secrets lie behind the intense gaze.
Prompt
facial-expressions Disgust: Disdain and disgust ; A hero, their face hardened with disgust, as they confront a corrupt politician; eye-level; Hero; a grand, opulent office with a powerful politician sitting behind a large desk; cinematic
Characteristic
Shot : A man in a suit is looking directly at the camera, he is in a room with a chandelier and candles in the background. Another man in a suit is standing behind him and to the right.
Aesthetic Score : 0.6
Mood : serious, intense, formal
Quality
Entropy : 6.52
Noise : 52
Prompt Clip Score : 0.27
AI Evaluation
Likelihood of AI : 0.20
Image errors : There is a slight blurring effect in the edges of the image, suggesting it has been processed or cropped. This is more noticeable around the edges of the man on the right.
Shadows and Secrets: A Haunting Walk Down a Dark Alley
A shadowy figure disappears into the depths of a dimly lit alleyway, leaving behind an atmosphere of eerie suspense. The scene, filled with garbage bins and flickering streetlights, evokes a sense of mystery and intrigue, leaving you wondering what secrets lie hidden in the darkness.
Prompt
facial-expressions Disgust: Despair and alienation ; A lone figure, hunched over in a dimly lit alleyway; eye-level; Single Person; overflowing trash bins and graffiti-covered walls; cinematic
Characteristic
Shot : A hooded figure stands in a dark alleyway, illuminated by street lights. The alley is lined with garbage bins on either side.
Aesthetic Score : 0.4
Mood : dark, mysterious, suspenseful
Quality
Entropy : 6.36
Noise : 88
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : The image appears to be slightly overexposed in some areas, which makes it difficult to see detail.
Superman’s Determined Gaze: A Moment of Grit in the Midst of Battle
A close-up shot captures Superman’s intense expression, his face etched with determination as he faces an unseen foe. Bloodstains on his costume hint at the intensity of the fight, creating a sense of drama and urgency. The mood is powerful and intense, leaving the viewer on the edge of their seat.
Prompt
facial-expressions Disgust: Horror and disgust ; A superhero, their face contorted in revulsion, as they witness a horrific crime; eye-level; Hero; a chaotic crime scene with blood and debris; cinematic
Characteristic
Shot : A close-up portrait of a man dressed as Superman, his face contorted in a furious expression, with blood smeared on his face and costume.
Aesthetic Score : 0.6
Mood : intense, aggressive, dramatic
Quality
Entropy : 6.64
Noise : 88
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.10
Image errors : There are no visible artifacts or errors in the image.
Intrigue in the Shadows: A Man in Red, a City of Secrets
A man in a striking red suit stands amidst the urban landscape, his gaze fixed on something unseen. The atmosphere is thick with mystery, enhanced by the dramatic lighting and the blurred figure of another man disappearing into the night. Is this a moment from a film, or a glimpse into a world of secrets?
Prompt
facial-expressions Disgust: Anger and disgust ; A superhero, their face etched with disgust, as they confront a villain who has committed a heinous act; eye-level; Hero; a dark and smoky cityscape with a towering villainous figure; cinematic
Characteristic
Shot : A man in a red and gold costume is walking in a city. He looks angry and determined. The city is dark and mysterious, but it also has a sense of hope and possibility.
Aesthetic Score : 0.6
Mood : dark, determined, hopeful
Quality
Entropy : 6.37
Noise : 58
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : The image has some artifacts. There are some pixels that are slightly out of place, and the colors are a bit too saturated. However, these errors are not very noticeable.
Lost in the Digital Shadows: A Man Confronts the Unseen
A chilling image captures a man immersed in a dark room, headphones on, eyes fixed on a computer screen. The screen displays a spectral figure, adding to the eerie atmosphere. Shadows and low light create a sense of mystery and suspense, leaving the viewer to wonder what secrets lie within the digital realm.
Prompt
facial-expressions Disgust: Fear and disgust ; A gamer, their face pale and sweaty, as they witness a disturbing scene in a horror game; eye-level; Gamer; a dimly lit gaming room with a flickering monitor and a dark, ominous atmosphere; cinematic
Characteristic
Shot : A man wearing headphones is sitting in front of a computer screen. The screen is displaying a scene of a ghost or shadowy figure. The room is dimly lit and the man’s face is partially obscured by shadows. He looks focused and perhaps slightly frightened.
Aesthetic Score : 0.6
Mood : mysterious, suspenseful, dark
Quality
Entropy : 5.80
Noise : 46
Prompt Clip Score : 0.24
AI Evaluation
Likelihood of AI : 0.10
Image errors : None
French Fry Faux Pas: Woman’s Dinner Interrupted by Unwelcome Guest
A woman’s dining experience takes a turn for the worse when she discovers a cockroach nestled amongst her french fries. The image captures her disgust and the surreal nature of the unexpected encounter.
Prompt
facial-expressions Disgust: Revulsion and disgust ; A woman, her face contorted in disgust, as she discovers a cockroach in her food; eye-level; Single Person; a brightly lit restaurant with a table full of food and a cockroach crawling on a plate; cinematic
Characteristic
Shot : A woman is looking at a plate of fries with a large insect on top. She is sitting at a table in a restaurant.
Aesthetic Score : 0.3
Mood : surprised, confused, disgust
Quality
Entropy : 6.73
Noise : 66
Prompt Clip Score : 0.31
AI Evaluation
Likelihood of AI : 0.10
Image errors : The image is slightly out of focus and the lighting is uneven.
Lost in the Digital Abyss: A Face of Fear and Wonder
A VR headset user is captivated by a monstrous face on their computer screen, creating a scene of intense, futuristic surrealism. The image evokes a sense of mystery and excitement, leaving the viewer wondering what lies beyond the digital veil.
Prompt
facial-expressions Disgust: Unease and disgust ; A gamer, their eyes wide with disgust, as they witness a grotesque scene in a virtual reality game; eye-level; Gamer; a brightly lit gaming room with multiple monitors and controllers; cinematic
Characteristic
Shot : A person wearing a VR headset is sitting in a dimly lit room, using a computer. The person is looking at the VR screen, which is displaying a virtual environment. There is a monitor in the background, displaying a graphic image, possibly from the virtual environment.
Aesthetic Score : 0.7
Mood : focused, futuristic, immersive
Quality
Entropy : 6.43
Noise : 60
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : There is a slight color banding issue in the background monitor.
Innocence Confronted: A Girl, a Rat, and a Silent Witness
A chilling scene unfolds as a young girl kneels beside a dead rat in a dimly lit kitchen. Three adults stand silently in the background, their expressions unreadable. The image evokes a sense of unease and suspense, leaving the viewer to ponder the story behind this unsettling tableau.
Prompt
facial-expressions Disgust: Horror and disgust ; A family, their faces twisted in disgust, as they discover a dead rat in their kitchen; eye-level; Normal People; a cluttered and messy kitchen with dirty dishes and a overflowing trash can; cinematic
Characteristic
Shot : A young girl is kneeling on the floor in a kitchen. There are three other people in the background, two of whom are adults, and one of whom is a teen or young adult. The girl looks scared or worried. There is a dead rat on the floor.
Aesthetic Score : 0.5
Mood : suspense, unsettling, dark
Quality
Entropy : 6.83
Noise : 87
Prompt Clip Score : 0.31
AI Evaluation
Likelihood of AI : 0.20
Image errors : There are some minor image errors, such as noise and artifacts. Some colors seem to be oversaturated.
A Moment of Reflection
A close-up shot captures an older woman lost in thought as she contemplates a plate of food. The lighting and composition evoke a sense of melancholy and loneliness, highlighting the introspective nature of the moment.
Prompt
facial-expressions Disgust: Disappointment and disgust ; A young woman, her face pale and wrinkled, as she stares at a plate of spoiled food; eye-level; Normal Person; a cluttered kitchen with dirty dishes and a overflowing trash can; cinematic
Characteristic
Shot : A woman sitting at a table in a kitchen, looking directly at the camera, with a plate of food in front of her. There is a kitchen sink behind her, and a white counter in front of her.
Aesthetic Score : 0.4
Mood : pensive, lonely, melancholic
Quality
Entropy : 6.90
Noise : 74
Prompt Clip Score : 0.31
AI Evaluation
Likelihood of AI : 0.10
Image errors : The image is slightly overexposed. The woman’s skin appears slightly too bright. There are some artifacts in the background, likely due to noise or compression.
Lost in the Urban Wasteland
A solitary figure, clad in a dark green jacket and baseball cap, walks through a desolate street littered with garbage cans and debris. The scene evokes a sense of gloom and isolation, highlighting the harsh realities of urban life.
Prompt
facial-expressions Disgust: Repulsion and disgust ; A man, his face contorted in disgust, as he walks past a pile of rotting garbage; eye-level; Single Person; a dirty and neglected street with overflowing trash cans; cinematic
Characteristic
Shot : A man in a dark green jacket and jeans walks down a narrow street, lined with trash bins on both sides, with buildings on either side.
Aesthetic Score : 0.5
Mood : urban, dreary, contemplative
Quality
Entropy : 6.84
Noise : 93
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.10
Image errors : The image is slightly overexposed, and the colors are a bit muted. Some areas are soft and blurry, but this could be intended.
Conclusion
The analysis shows that the generative AI model performed well in understanding the camera position and shot composition, but struggled with the aesthetic expectations. Here’s a breakdown:
- Camera Position: The model scored 0.35, which is considered below average. This suggests that the generated image didn’t accurately reflect the camera position described in the prompt.
- Shot Analysis: The model scored 0.63, which is considered good. This indicates that the model was able to understand the scene and create a shot that was generally consistent with the prompt.
- Aesthetic Analysis: The model scored 0.17, which is considered below average. This means that the generated image didn’t match the expected aesthetic style as closely as it could have.
Overall, the model demonstrated a good understanding of shot composition but struggled with camera positioning and aesthetic expectations.
Sources:
- https://dramaresource.com/storytelling/
- https://seedsoftellers.eu/resources/the-body-language-for-young-tellers/
- https://digitalcollections.sit.edu/cgi/viewcontent.cgi?article=1288&context=sandanona&filename=1&type=additional
- https://citeseerx.ist.psu.edu/document?doi=7f842882e9bb1fa2c0e96939bc8d2c37e34e17c0&repid=rep1&type=pdf
- https://www.twinkl.co.uk/search?q=drama+facial+expression
- https://fal.ai/models/fal-ai/flux/dev/api