AI's Facial Expressions: A Mixed Bag of Success with Stable-diffusion
- 9 minutes read - 1774 wordsTable of Contents
Facial expressions are a powerful tool for conveying emotions and adding depth to visual storytelling. In the realm of generative AI, the ability to create images with realistic and nuanced facial expressions is a crucial aspect of achieving lifelike and engaging visuals. This blog post explores the capabilities of a generative AI model in capturing facial expressions, analyzing its performance across various scenarios and highlighting its strengths and weaknesses.
Created with: stability-ai-core
Lost in the City Lights
A solitary figure shrouded in darkness stands against a backdrop of vibrant city lights, creating a mood of mystery and melancholic intrigue. The blurred background suggests a sense of isolation and the woman’s expression hints at a story waiting to be told.
Prompt
facial-expressions Agreement: melancholy, contemplative ; A lone figure; eye-level; Single Person; a bustling city street at night; cinematic
Characteristic
Shot : A woman stands in a city street at night, looking directly at the camera, with her face illuminated by streetlights.
Aesthetic Score : 0.7
Mood : mysterious, urban, pensive
Quality
Entropy : 6.18
Noise : 70
Prompt Clip Score : 0.23
AI Evaluation
Likelihood of AI : 0.10
Image errors : Minor noise in the darker areas of the image and slight blurriness on the woman’s face.
Superman Amidst the Ruins: A Hero’s Stand in a World of Destruction
A powerful image captures the essence of heroism amidst devastation. Superman, clad in his iconic suit, stands tall against a backdrop of burning buildings and billowing smoke. The juxtaposition of his unwavering presence with the surrounding chaos creates a dramatic and poignant scene, highlighting the weight of his responsibility in a world in need.
Prompt
facial-expressions Agreement: determined, resolute ; A superhero standing tall; eye-level; Hero; a cityscape with a burning building in the background; cinematic
Characteristic
Shot : A man dressed as Superman stands in front of a burning city. He is looking directly at the camera with a serious expression.
Aesthetic Score : 0.6
Mood : dramatic, heroic, intense
Quality
Entropy : 6.86
Noise : 79
Prompt Clip Score : 0.28
AI Evaluation
Likelihood of AI : 0.50
Image errors : The fire appears somewhat artificial and does not look very realistic.
The Joy of Family Gatherings
A heartwarming scene of a family sharing a meal together, radiating warmth, intimacy, and contentment. The soft lighting and comfortable atmosphere create a sense of togetherness and happiness, capturing the essence of a fulfilling moment.
Prompt
facial-expressions Agreement: peaceful, content ; A family gathered around a dinner table; eye-level; Normal People; a cozy kitchen with warm lighting; cinematic
Characteristic
Shot : A family gathered around a dinner table in a warmly lit kitchen, enjoying a meal. The table is set with food, wine glasses, and silverware, and there is a relaxed and convivial atmosphere.
Aesthetic Score : 0.7
Mood : warm, cozy, happy
Quality
Entropy : 6.74
Noise : 78
Prompt Clip Score : 0.27
AI Evaluation
Likelihood of AI : 0.20
Image errors : Some minor noise is visible in the background and shadows, particularly around the cabinets.
Neon Glow, Focused Flow: Gamer Immersed in Virtual World
A young man, headphones on, sits before a computer bathed in vibrant neon light. His intense focus suggests a thrilling game session, the futuristic atmosphere heightened by the dramatic lighting.
Prompt
facial-expressions Agreement: excited, engaged ; A gamer intensely focused on a screen; eye-level; Gamer; a dimly lit room with neon lights reflecting on the screen; cinematic
Characteristic
Shot : A young man wearing headphones is sitting in a gaming chair, typing on a keyboard. There are neon lights in the background.
Aesthetic Score : 0.6
Mood : focused, intense, futuristic
Quality
Entropy : 6.18
Noise : 64
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : There are no obvious artifacts or errors in the image.
Lost in Time: A Woman Walks Through a City’s Melancholy Past
A solitary figure in a black coat traverses a narrow cobblestone street, surrounded by aged brick buildings and fading grandeur. The scene evokes a sense of melancholic introspection, capturing the essence of a city’s forgotten history.
Prompt
facial-expressions Agreement: reflective, introspective ; A woman walking down a quiet street; eye-level; Single Person; a row of old, brick buildings with faded paint; cinematic
Characteristic
Shot : A woman in a black coat stands on a cobblestone street in an old city. The street is lined with brick buildings and there is a lamppost in the background. The photo has a vintage, film noir feel.
Aesthetic Score : 0.6
Mood : melancholy, atmospheric, mysterious
Quality
Entropy : 6.80
Noise : 82
Prompt Clip Score : 0.28
AI Evaluation
Likelihood of AI : 0.10
Image errors : There are some minor artifacts in the image, most notably in the background near the woman’s left shoulder. The image is also slightly overexposed, which might be a deliberate stylistic choice, but it could be improved with some editing.
Man Defies the Storm
A lone figure in a leather jacket stands resolute against a backdrop of a raging storm, lightning illuminating the dramatic scene. The contrast between his confident stance and the tempestuous sky creates a powerful and intense mood.
Prompt
facial-expressions Agreement: powerful, defiant ; A hero raising their fist in defiance; eye-level; Hero; a dark, stormy sky with lightning flashing in the background; cinematic
Characteristic
Shot : A man in a leather jacket stands in front of a stormy sky with lightning strikes in the background.
Aesthetic Score : 0.6
Mood : dramatic, intense, powerful
Quality
Entropy : 6.70
Noise : 65
Prompt Clip Score : 0.24
AI Evaluation
Likelihood of AI : 0.70
Image errors : The lightning strikes appear to be somewhat artificial and the man’s face seems slightly out of focus.
Laughter and Joy in the Park
A group of young adults share a moment of pure joy and laughter in a vibrant park setting. Captured from a low angle, the image offers an intimate glimpse into their carefree happiness.
Prompt
facial-expressions Agreement: joyful, carefree ; A group of friends laughing together; eye-level; Normal People; a sunny park with trees and flowers; cinematic
Characteristic
Shot : A group of friends are laughing together in a park. They are looking at the camera.
Aesthetic Score : 0.8
Mood : joyful, happy, carefree
Quality
Entropy : 6.86
Noise : 83
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : None
Victory Dance! Confetti Rain Down on Gamer’s Triumph
A young man celebrates a hard-earned victory in front of his computer, surrounded by a flurry of confetti. His raised arms and beaming smile capture the pure joy and energy of the moment.
Prompt
facial-expressions Agreement: triumphant, ecstatic ; A gamer celebrating a victory; eye-level; Gamer; a brightly lit room with confetti and streamers; cinematic
Characteristic
Shot : A young man wearing headphones is sitting at a computer desk celebrating with confetti. He is wearing a blue jacket with a ‘Prime Gamer’ t-shirt and is looking up towards the confetti as he raises his arms in celebration.
Aesthetic Score : 0.7
Mood : joyful, celebratory, energetic
Quality
Entropy : 6.64
Noise : 68
Prompt Clip Score : 0.29
AI Evaluation
Likelihood of AI : 0.20
Image errors : The image appears to be slightly overexposed, especially in the area of the confetti. There is also some noise visible in the shadows.
Autumnal Melancholy: A Man Finds Solitude in the Falling Leaves
A poignant image captures the essence of autumn, with a man lost in thought on a park bench beneath a canopy of yellowing leaves. The muted colors and his contemplative pose evoke a sense of quiet introspection and the bittersweet beauty of the season.
Prompt
facial-expressions Agreement: lonely, melancholic ; A man sitting alone on a bench; eye-level; Single Person; a deserted park with fallen leaves; cinematic
Characteristic
Shot : A man is sitting on a bench in a park with fall leaves on the ground. He is looking down and appears contemplative. The background is a blurred view of trees with fall foliage. There is a bench behind him.
Aesthetic Score : 0.7
Mood : melancholy, contemplative, autumnal
Quality
Entropy : 6.76
Noise : 71
Prompt Clip Score : 0.30
AI Evaluation
Likelihood of AI : 0.10
Image errors : No noticeable errors
Lost in the City Lights: A Moment of Contemplation
A solitary figure stands on a rooftop, silhouetted against the vibrant cityscape. The night sky, a canvas of deep blue, reflects the man’s introspective mood as he gazes out at the distant lights. The scene evokes a sense of nostalgia and urban solitude, leaving the viewer to ponder the man’s thoughts and the vastness of the city around him.
Prompt
facial-expressions Agreement: determined, hopeful ; A hero standing on a rooftop overlooking the city; eye-level; Hero; a panoramic view of a city skyline at night; cinematic
Characteristic
Shot : A man is standing on a rooftop overlooking a city at dusk. The city lights are glowing in the distance. The man is wearing a jacket and jeans. The image is a bit dark, but the colors are good.
Aesthetic Score : 0.7
Mood : reflective, contemplative, urban
Quality
Entropy : 6.60
Noise : 70
Prompt Clip Score : 0.23
AI Evaluation
Likelihood of AI : 0.10
Image errors : There is some slight noise in the image, particularly in the darker areas.
Conclusion
The results show that the generative AI model performed well in understanding the scene and camera position, but struggled with the aesthetic aspect. Here’s a breakdown:
- Camera Position: The model scored 0.1, indicating it did not perform well in capturing the intended camera position. This suggests the model may not be very sensitive to camera position instructions.
- Shot Analysis: The model scored 0.36, indicating it performed moderately well in understanding the scene described in the prompt. This suggests the model can grasp some aspects of the scene but may not be able to fully capture the intended shot.
- Aesthetic Analysis: The model scored 0.06, indicating it performed very well in capturing the intended aesthetic. This suggests the model is able to generate images that closely match the desired aesthetic style.
Overall, the model shows promise in understanding the scene and capturing the desired aesthetic, but needs improvement in accurately interpreting camera position instructions.
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://stability.ai