sjs/gh-1SecondEveryday-image-analysis-eval
1
0
Fork 0
mirror of https://github.com/1SecondEveryday/image-analysis-eval.git synced 2026-03-25 09:05:49 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Remove under-performing prompts

Browse source
This commit is contained in:
Sami Samhuri 2025-06-25 00:29:55 -04:00
parent dd16b71f54
commit e706d68e72
No known key found for this signature in database
8 changed files with 0 additions and 90 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
prompts/02-scene-everything.txt
View file

@ -1 +0,0 @@
Output only comma-separated keywords that cover everything in the scene: colors, objects, people (if present, use 'people' with descriptive count like '2-people'), what they're doing and feeling, the environment (location type, weather, lighting, time), and the general mood or vibe.

1
prompts/04-generate-all-aspects.txt
View file

@ -1 +0,0 @@
Generate comma-separated keywords only: include all colors, objects, number of people, actions, facial expressions or emotions, the environment (location type, weather, lighting, time), and the scene's mood or atmosphere.

1
prompts/06-concise-complete.txt
View file

@ -1 +0,0 @@
Comma-separated keywords covering: colors, all objects, people (if humans visible, include 'people' with count like '3-people'), activities, emotions, location type, weather/lighting, time of day, mood. Keywords only.

21
prompts/07-ultra-detailed-scene.txt
View file

@ -1,21 +0,0 @@
Generate comprehensive search keywords for this image as comma-separated values. Analyze and keyword:
PEOPLE: if humans are visible include 'people' keyword followed by descriptive count (e.g. '3-people', 'group', 'couple'), apparent ages (baby/child/teen/adult/elderly), genders if clear, relationships (couple/family/friends/strangers), body language, facial expressions (smiling/laughing/crying/serious/surprised), activities (eating/walking/sitting/playing/working). If no people are present, skip all people-related keywords
MOOD & ATMOSPHERE: overall emotional tone (joyful/peaceful/tense/romantic/nostalgic/energetic), energy level (calm/lively/chaotic), formality (casual/formal/ceremonial)
SETTING: indoor/outdoor, specific location type (beach/mountain/city/restaurant/home/office/park), country/region if identifiable, venue type (public/private)
TIME & LIGHTING: time of day (early morning/morning/midday/afternoon/golden hour/evening/night), lighting quality (bright/soft/harsh/dim/dramatic), light source (natural/artificial/mixed)
COLORS: dominant colors (top 3), accent colors, color mood (warm/cool/neutral/vibrant/muted)
OBJECTS: food types if visible (cuisine type, dishes, drinks), vehicles, furniture, technology, decorations, sports equipment, musical instruments, art
NATURE: weather conditions, season indicators, landscapes, water bodies, vegetation, animals
BACKGROUND: what's behind the main subjects, architectural elements, crowds, signage
EVENTS: if applicable (wedding/birthday/concert/sports/holiday/vacation)
Output only comma-separated keywords, no explanations.

13
prompts/08-memory-search-optimizer.txt
View file

@ -1,13 +0,0 @@
Create detailed keywords for video diary search. Users might search for: "happy moments", "food experiences", "family time", "adventures", "quiet moments", "celebrations", "daily life", "travel memories".
Keyword everything visible including:
- People: count, approximate ages, emotions on faces, what they're doing, how they're interacting
- Scene type: where this is happening, indoor/outdoor, public/private space
- Time: morning light, afternoon, golden hour, evening, night time
- Mood: the feeling of the moment (joyful, peaceful, exciting, intimate, festive, contemplative)
- Activities: eating, playing, working, relaxing, traveling, celebrating, exploring
- Details: specific foods visible, drinks, decorations, clothing styles, weather, season
- Colors: main colors that define the scene
- Special moments: laughter, hugs, cheers, surprises, achievements
Format: comma-separated keywords only, be specific rather than generic.

19
prompts/09-contextual-story-tagger.txt
View file

@ -1,19 +0,0 @@
Analyze this image as a moment in someone's life story. Generate specific, searchable keywords covering:
WHO: number of people, their likely relationships, emotions showing on faces, age groups
WHAT: main activity, secondary activities, interactions between people, gestures
WHERE: type of location, indoor/outdoor, specific venue (restaurant/home/beach/etc), geographic hints
WHEN: time of day, season clues, weather conditions, lighting quality
WHY: occasion if apparent (meal/celebration/vacation/work/leisure), mood of gathering
Include sensory details:
- Visual: dominant colors, lighting (harsh/soft/golden), contrast
- Implied: likely sounds (quiet/loud/music), temperature (hot/cold), atmosphere
Keyword specific items:
- Food: cuisine type, specific dishes if identifiable, drinks
- Objects: technology, vehicles, sports equipment, decorations
- Nature: landscapes, water, sky conditions, plants, animals
- Clothing: formal/casual, weather-appropriate
Output format: detailed comma-separated keywords only.

21
prompts/10-moment-finder-pro.txt
View file

@ -1,21 +0,0 @@
Generate exhaustive search keywords. Think like someone searching their memories:
EMOTION KEYWORDS: happy, laughing, smiling, serious, contemplative, excited, surprised, loving, playful, tired, focused, relaxed, celebratory
PEOPLE KEYWORDS: exact count, "group of friends", "couple", "family", "solo", ages if clear, interactions
ACTIVITY KEYWORDS: eating, drinking, cooking, playing, sports, reading, working, dancing, hugging, talking, walking, sitting, lying down
PLACE KEYWORDS: home, restaurant, cafe, beach, mountain, city, countryside, indoor, outdoor, kitchen, living room, bedroom, office, street, park, venue name if visible
TIME KEYWORDS: sunrise, morning, midday, afternoon, sunset, golden hour, blue hour, evening, night, specific time if visible
OBJECT KEYWORDS: list everything visible - food items, drinks, furniture, electronics, vehicles, decorations, plants, books, instruments
ATMOSPHERE KEYWORDS: cozy, energetic, romantic, professional, casual, festive, quiet, busy, crowded, intimate
COLOR KEYWORDS: list prominent colors
WEATHER/SEASON KEYWORDS: sunny, cloudy, rainy, snowy, fog, clear, spring, summer, fall, winter indicators
Comma-separated only, no duplicates.

13
prompts/11-smart-scene-decoder.txt
View file

@ -1,13 +0,0 @@
You're analyzing frames for an AI-powered video diary search. Users search with natural language like "dinner with friends", "kids playing", "sunset at the beach", "birthday celebrations", "quiet morning coffee".
Extract and keyword:
HUMANS: if humans visible include 'people' and descriptive count (e.g. '4-people', 'couple', 'crowd'), estimated ages in decades (20s/30s/etc), primary emotion per person, body language, attire style. Skip if no humans present
ACTIONS: primary action, secondary actions, interactions, gestures
LOCATION: venue type, indoor/outdoor, architectural style, geographic region if evident
TEMPORAL: exact time if visible, otherwise: dawn/morning/noon/afternoon/dusk/night, season indicators
AMBIANCE: energy level 1-10, mood descriptors, lighting quality, color temperature
OBJECTS: enumerate all significant objects, food with cuisine type, beverages, decorations
CONTEXT: occasion type, relationship dynamics, cultural indicators
TECHNICAL: image quality descriptors, composition style
Output as comma-separated keywords. Prioritize specific over generic (e.g., "pepperoni pizza" not just "food").