Upload 16 files

.gitattributes

@@ -33,3 +33,10 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/figures/framework.jpg filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-1/reference_image.png filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-2/reference_image.png filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-3/reference_image.png filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-4/reference_image.png filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-5/reference_image.png filter=lfs diff=lfs merge=lfs -text
+assets/ti2va_cases/case-6/reference_image.png filter=lfs diff=lfs merge=lfs -text

assets/ti2va_cases/case-1/prompt.txt

@@ -0,0 +1 @@
+No background music, quiet recording environment, close-mic capture, no noticeable reverb. Close-up, eye-level view, softly lit with frontal lighting. The frame features an elderly white woman with short gray curly hair, blue eyes, deeply lined face, flushed cheeks, and red lipstick. She wears a black coat and a red scarf with a white floral pattern, a black microphone clipped to her collar. She sits in front of a brown wooden piece of furniture, with a pale yellow textured wall behind her, on which hangs a miniature model depicting a canal and Dutch-style buildings on both sides. Initially, the elderly white woman faces the camera, lips moving, speaking in a hoarse voice with a heavy expression<speech>weg haden ze nog gezegd, ze liepen weg.</speech>. Then, she briefly lowers her head, her gaze moving downward, mouth slightly open as if sighing. Immediately afterward, she raises her head to look at the camera again, continuing<speech>Nou, dan kwam de directie er bij, want de directeur moest oepen.</speech>, blinking slowly throughout, her expression solemn.

assets/ti2va_cases/case-2/prompt.txt

@@ -0,0 +1 @@
+Continuous <sfx>ambient store sounds, including people talking and footsteps</sfx> throughout. Medium close-up, eye-level view. Inside an indoor store with bright lighting. The main focus is a man and a woman conversing. On the left side of the frame is a young-to-middle-aged white woman wearing glasses, with brown wavy medium-short hair, wearing a black jacket and a dark red inner layer, holding a dark-colored wine bottle with a black label featuring gold English text <text>THE髭振酒</text>. On the right side is a middle-aged white man with a beard, receding hairline, wearing a gray plaid suit jacket and a light blue shirt, holding a green wine bottle with a white label featuring a minimalist black line pattern and text. Behind them are shelves filled with products, with multiple customers and staff walking and shopping in the background, and a row of track spotlights on the ceiling. Initially, the man in the gray suit holds up the green wine bottle, looking at the woman across from him. Then he dramatically widens his eyes, showing the bottle to her, enthusiastically saying <speech>And the second one that we have a wine.</speech>. The woman lowers her head, looking at her own dark-colored bottle, smiling. Immediately afterward, the woman in the black jacket looks up at the man, asking curiously <speech>They are also do a wine?</speech>. The man nods affirmatively, saying <speech>Yep.</speech>. The woman confirms, saying <speech>The messa.</speech>. The man then raises his left hand in a gesture of explanation, continuing to speak <speech>So when you don't know it, you're kind of afraid because maple.</speech>"}

assets/ti2va_cases/case-3/prompt.txt

@@ -0,0 +1 @@
+Continuous <music>solemn and expressive classical chamber music performance throughout, featuring two violins, one cello, one harpsichord, and one bassoon, with intertwined melodies and rich harmonies producing noticeable reverb in the stone-walled church space.</music>. High-angle wide shot, fixed frame: Gothic stone church interior. Soft, cool light from left arched window. White walls with carved memorials and shield emblems. Gray stone floor, red-and-black diamond tile path center. Eight musicians in semicircle. Center: young white male harpsichordist (black long-sleeve, facing away). Foreground left: middle-aged white female violinist (blonde, purple lace top, black dress). Beside her: balding middle-aged white male violinist. Foreground right: young white male cellist (brown short hair). Background: middle-aged white female violinist (blonde), balding middle-aged white male bassoonist, two standing male musicians. Each has black music stand. Microphones on floor. As the shot begins, all musicians start playing their instruments, gently swaying their bodies in rhythm with the music. <music>A solemn and expressive classical chamber music performance begins, featuring two violins, one cello, one harpsichord, and one bassoon, with intertwined melodies and rich harmonies producing noticeable reverb in the stone-walled church space.</music> The male and female violinists on the left and the cellist on the right are focusedly drawing their bows across the strings, fingers moving swiftly along the strings. The central harpsichordist's hands move rapidly across the keyboard while glancing sideways at the sheet music. The bassoonist in the background is playing his instrument. As the piece progresses, a cellist from the far right edge of the frame enters the shot and begins playing.

assets/ti2va_cases/case-4/prompt.txt

@@ -0,0 +1 @@
+<sfx>Constant noisy background sounds from the gym throughout, including equipment clanging, dumbbell plates hitting, and muffled conversations, with noticeable spatial reverb.</sfx>. Medium shot with a slight high-angle view, indoor gym setting, bright lighting. The main subject is a young white woman sitting on a black gym bench, with long golden ponytail, wearing white in-ear headphones, and a large black mandala tattoo on her left arm, though details like the medical giant logo beneath are somewhat blurred. She wears a white sports bra and olive-green ribbed leggings, and black-and-white Adidas sneakers. On the left side of the frame is a blue fitness equipment column with a warning label, an orange warning tag marked <text>WARNING</text>, and a white label below. The background is a white wall with blue brand text visible, partially showing <text>Nordic</text>, and <text>to</text> visible on the right. Other gym equipment and multiple members exercising are visible, including a man in a black hoodie and a woman in a light gray hoodie. Initially, the blonde woman keeps her upper body lifted with bent legs, arms hugging her knees close to her chest. Her mouth is slightly open, looking directly at the camera, beginning her explanation in a slightly hoarse but clear and strong voice in Swedish: <speech>So try to do or this can a try to move bengal。</speech>. Then, while maintaining the knee-hug position, she extends her right index finger forward, then brings it back, adjusting her left hand's position on her knee. Next, she moves her left hand to support her left hip and waist, while bringing her right knee to the side and tilting her body backward. Subsequently, she keeps her left leg bent for support, extends her right leg, and slowly moves it upward and downward in rhythm with the command, completing one snow angel-like motion. Throughout and after the movement, she continues speaking directly to the camera: <speech>Can she tell me about the room and lift it little from that? Are they just broke in them now we could。</speech>

assets/ti2va_cases/case-5/prompt.txt

@@ -0,0 +1 @@
+Continuous <sfx>sound of waves and water currents around the boat</sfx> throughout. Medium shot, eye-level view. During the day, in clear outdoor conditions, a middle-aged white man stands on a boat facing the camera. He wears a gray baseball cap and blue-tinted sunglasses, a gray short-sleeved T-shirt with a brand logo featuring <text>YETI</text> on the chest, mostly obscured by the large fish he holds. He cradles a massive dark brown grouper with a wet, shiny surface reflecting sunlight and a serrated dorsal fin. A white outboard motor is located behind his left, marked with black <text>EVINRUDE</text>, partially hidden by the fish's tail at times. On the right side of the frame, another man's profile is visible—bearded, wearing a hat, holding a fishing rod. The background shows a vast blue ocean under a sky with a few scattered clouds. Initially, the man with sunglasses beams a bright smile, enthusiastically saying <speech>I mean, so far we've caught Marlin, Cooper, I mean, a Wahoo，we got slam. Why stop now？</speech> to the camera. His expression is highly animated. Then, he briefly turns his head to the right to look at the man with the fishing rod beside him, before turning back, maintaining a joyful expression throughout. The camera gently shakes due to the boat's movement during the entire sequence.

assets/ti2va_cases/case-6/prompt.txt

@@ -0,0 +1 @@
+No background music, with slight ambient outdoor noise. Eye-level medium close-up, centered composition. On a sunny day, a young white woman and a young South Asian man sit at a round black grid table under a black patio umbrella. The woman is on the left side of the frame, with long brown hair, wearing a pink sleeveless dress, and a thin black bracelet on her left wrist. The man is on the right side, with short black hair, wearing a blue short-sleeve Polo shirt with a black microphone clipped at the collar. A glass of water sits on the table in front of them. The scene is set on a wooden deck, with open water, distant apartment buildings, and green trees in the background, and large white clouds in the sky. Initially, the man in the blue Polo shirt looks at the woman across from him, raises both hands in gesture, and says in a gentle tone <speech>That's okay. We can order。</speech>. Then, the woman in the pink dress becomes serious, lowers her hands, and gently shakes her head, responding with a disappointed tone <speech>It's not about the wine. Nothing ever works out for me.</speech>. She then lifts her right hand and touches the glass on the table. The man leans forward slightly, listening intently. The woman then asks in a curious, probing tone <speech>Why do you say that?</speech>. The woman continues to look at him seriously and explains <speech>You never smelled at me when I smelled at you back at the start, so I decided to use hope this time.</speech>