AI Captures the Essence of Emotion, But Struggles with Camera Angles with Titan-g1
- 9 minutes read - 1873 wordsTable of Contents
Facial expressions are a powerful tool for conveying emotions and intentions. In the realm of generative AI, the ability to create realistic and expressive faces is a crucial benchmark. This blog post explores the performance of a generative AI model in generating facial expressions across a range of scenes, analyzing its strengths and weaknesses in capturing camera position, scene context, and aesthetic elements. We’ll delve into specific examples, highlighting how the model excels in certain areas while demonstrating room for improvement in others. Join us as we explore the fascinating world of AI-generated facial expressions and their potential to enhance storytelling and visual communication.
Created with: titan-g1
Joyful Laughter and Animated Gestures
A person with short dark hair, wearing a dark blue t-shirt, bursts into laughter and animated gestures against a plain gray wall. Their energy and vitality are palpable, capturing a moment of pure joy.
Prompt
facial-expressions Embarrassment: Awkward and self-conscious ; A single woman; eye-level; Single Persons; A crowded cafe with loud chatter and laughter; cinematic
Characteristic
Shot : A person, likely a woman, is laughing and talking with their hands raised in a gesture of excitement or enthusiasm. The background is out of focus, suggesting an indoor setting.
Aesthetic Score : 0.7
Mood : joyful, lively, positive
Quality
Entropy : 6.70
Noise : 94
Prompt Clip Score : 0.15
AI Evaluation
Likelihood of AI : 0.10
Image errors : No noticeable errors.
A Burst of Color and Energy: Man in Red and Gold Dances at Festive Celebration
A man in a vibrant red and gold costume, adorned with a crown, takes center stage in a bustling street. His energetic dance moves command attention, while the blurred background hints at the lively atmosphere of a cultural festival or celebration. The scene exudes a festive mood, capturing the joy and vibrancy of the event.
Prompt
facial-expressions Embarrassment: Humiliated and exposed ; A vibrant carnival performer in full costume, standing on a stage in the center of a bustling marketplace, with crowds of people watching in awe.; cinematic
Characteristic
Shot : A man dressed in a flamboyant costume, possibly for a carnival or parade, is walking towards the camera with his back turned. A crowd of people are in the background.
Aesthetic Score : 0.7
Mood : festive, joyous, celebratory
Quality
Entropy : 6.88
Noise : 105
Prompt Clip Score : 0.19
AI Evaluation
Likelihood of AI : 0.10
Image errors : The image has some minor noise and artifacts, especially in the background. There’s a slight blurriness in the subject’s face, which could be a technical issue.
A Tense Exchange at a Formal Gathering
A man in a tuxedo speaks urgently to a woman in a black dress at a formal event. The dim lighting and his hand gesture create a sense of tension, hinting at a secret or a difficult conversation. The presence of a third person in the background adds to the intrigue.
Prompt
facial-expressions Embarrassment: Mortified and ashamed ; A man in a business suit; eye-level; Normal People; A formal dinner party with elegant guests; cinematic
Characteristic
Shot : A man in a suit is talking to a woman in a black dress. They are both looking at each other. There is a glass of wine on the table in front of them.
Aesthetic Score : 0.6
Mood : serious, intimate, elegant
Quality
Entropy : 6.70
Noise : 100
Prompt Clip Score : 0.25
AI Evaluation
Likelihood of AI : 0.20
Image errors : There is a small amount of noise in the image, particularly in the shadows. The skin tones are a bit over-saturated, which makes the man’s skin look unnaturally red.
Victory Dance! Gamer Celebrates Epic Win
Capture the joy of a gamer’s triumph with this image. A young man, headphones on and hands raised in victory, sits in his gaming chair, pizza box in the foreground, radiating pure excitement and energy. The scene is full of playful energy and captures the thrill of a hard-fought win.
Prompt
facial-expressions Embarrassment: Cringing and defeated ; A gamer in a gaming chair; eye-level; Gamer; A dimly lit room with flashing screens and empty pizza boxes; cinematic
Characteristic
Shot : A young man wearing headphones is sitting in a gaming chair and is laughing. There is a pizza box in front of him on a desk.
Aesthetic Score : 0.6
Mood : joyful, playful, excited
Quality
Entropy : 6.85
Noise : 102
Prompt Clip Score : 0.26
AI Evaluation
Likelihood of AI : 0.20
Image errors : There are no major image errors, the background and subject appear naturally.
Bride’s Joyful Laughter Lights Up Wedding Celebration
A bride radiates happiness, clapping her hands and laughing with infectious joy, surrounded by her loved ones at a wedding. The image captures the pure delight of the moment, with the bride’s laughter taking center stage.
Prompt
facial-expressions Embarrassment: Lonely and out of place ; A woman in a wedding dress; eye-level; Single Persons; A crowded wedding reception with happy couples; cinematic
Characteristic
Shot : A bride is laughing with her friends at a wedding. The bride is wearing a white dress and a tiara. She is surrounded by her friends who are also wearing white dresses. The bride is the focal point of the image.
Aesthetic Score : 0.75
Mood : joyful, celebratory, romantic
Quality
Entropy : 6.57
Noise : 96
Prompt Clip Score : 0.17
AI Evaluation
Likelihood of AI : 0.10
Image errors : There are no visible errors in the image.
Superhero Victory: A Moment of Triumph Captured
This image captures the joy and confidence of a young woman dressed as a superhero, standing triumphantly in front of a modern building. Her raised arms and excited expression convey a sense of power and victory, while the blurred background adds to the dynamic energy of the scene.
Prompt
facial-expressions Embarrassment: Embarrassed and self-conscious ; A superhero in a cape; eye-level; Heroes; A cheering crowd at a victory parade; cinematic
Characteristic
Shot : A woman dressed as a superhero is standing in front of a building, raising her arms in victory. The scene is slightly blurred as if the photo is taken in motion.
Aesthetic Score : 0.7
Mood : joyful, empowered, hopeful
Quality
Entropy : 6.66
Noise : 97
Prompt Clip Score : 0.20
AI Evaluation
Likelihood of AI : 0.10
Image errors : There is slight blur in the background and the subject. This blur could be a result of the motion or camera settings.
Intense Conversation in a Warm Ambiance
A woman in a blazer sits at a table, her expression and gestures conveying urgency and excitement as she talks to someone off-camera. The warm, inviting restaurant setting and soft lighting create an intimate atmosphere, highlighting the intensity of the moment.
Prompt
facial-expressions Embarrassment: Uncomfortable and out of place ; A woman in a casual outfit; eye-level; Normal People; A fancy restaurant with white tablecloths and expensive wine; cinematic
Characteristic
Shot : A woman is sitting at a table in a restaurant, talking to someone off-camera. There’s a glass of wine on the table in front of her.
Aesthetic Score : 0.5
Mood : serious, engaged, conversational
Quality
Entropy : 6.59
Noise : 99
Prompt Clip Score : 0.24
AI Evaluation
Likelihood of AI : 0.10
Image errors : No visible errors or artifacts.
The Thrill of Victory: Gamer’s Intense Reaction Captured
A close-up shot reveals a young gamer’s intense excitement as they react to a thrilling moment in their game. The focused expression and dramatic lighting create a sense of tension and anticipation, capturing the raw emotion of competitive gaming.
Prompt
facial-expressions Embarrassment: Humiliated and defeated ; A gamer in a hoodie; eye-level; Gamer; A crowded esports tournament with loud cheers and flashing lights; cinematic
Characteristic
Shot : A young woman is sitting in a gaming chair, wearing a headset, and celebrating a win on her computer. She’s shouting and raising her fist in the air, with an intense expression. There is a person sitting behind her, and the scene is illuminated with blue and red light.
Aesthetic Score : 0.6
Mood : excited, intense, energetic
Quality
Entropy : 6.72
Noise : 108
Prompt Clip Score : 0.28
AI Evaluation
Likelihood of AI : 0.10
Image errors : No significant errors. The image is well-exposed and there is no noticeable noise.
A Moment of Reflection in Candlelight
A man in a tuxedo, bathed in the soft glow of candlelight, sits lost in thought. The scene evokes a sense of romance, elegance, and mystery, leaving the viewer to ponder the emotions behind his contemplative gaze.
Prompt
facial-expressions Embarrassment: Awkward and uncomfortable ; A man in a tuxedo; eye-level; Single Persons; A romantic dinner for two with candles and flowers; cinematic
Characteristic
Shot : A man in a suit and bow tie is sitting at a table with candles, looking thoughtful. The setting appears to be a romantic dinner date.
Aesthetic Score : 0.6
Mood : romantic, intimate, thoughtful
Quality
Entropy : 6.64
Noise : 99
Prompt Clip Score : 0.21
AI Evaluation
Likelihood of AI : 0.20
Image errors : The image appears to be slightly overexposed. There are some minor artifacts around the edges of the man’s suit.
Red-Clad Hero Caught in a Moment of Surprise
A man in a striking red cape and mask stands against a dark backdrop, his hands outstretched in a gesture of surprise. The dramatic pose and his wide-eyed expression create a humorous and slightly awkward scene, leaving us wondering what unexpected event has caught him off guard.
Prompt
facial-expressions Embarrassment: Mortified and ashamed ; A superhero in a mask; eye-level; Heroes; A news conference with reporters asking difficult questions; cinematic
Characteristic
Shot : A man dressed as a superhero, wearing a red cape and a red mask, is looking at the camera with a surprised expression.
Aesthetic Score : 0.3
Mood : surprised, comical, awkward
Quality
Entropy : 6.26
Noise : 96
Prompt Clip Score : 0.24
AI Evaluation
Likelihood of AI : 0.20
Image errors : The image appears to be slightly overexposed, leading to a lack of detail in the shadows.
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 0.2, which is considered below average. This suggests that the model didn’t accurately capture the intended camera position described in the prompt.
- Shot Analysis: The model scored 0.65, which is considered good. This indicates that the model was able to understand the scene described in the prompt and create a shot that aligns with it.
- Aesthetic Analysis: The model scored 0.12, which is considered very good. This means that the generated image’s aesthetic closely matched the expected aesthetic described in the prompt.
Overall, the model demonstrates a good understanding of the scene and its aesthetic, but needs improvement in accurately capturing the intended camera position.
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://docs.aws.amazon.com/bedrock/latest/userguide/titan-image-models.html