Group6 - Final Project Midterm Report

Progress So Far

The Effect

Storyline/Storyboard

Scene 1	Scene 2	Scene 3	Scene 4

• Scene 1

  • This scene will be much like the initial scene in the HP PictureBook Ad. It will consist of a group of people holding white frames, they will put the frame in front of their face and it will turn into a still image.
    • Number of Actors: 9
    • Video Length: ~5 sec

• Scene 2

  • Scene 2 will consists of a person in the foreground pulling pictures from the background. The background footage will be someone jumping up and down on a little trampoline; the person in the foreground will pull a picture from the scene that includes the trampoline which will result in the trampoline being removed from the scene; the person jumping will then fall to the ground.
    • Number of Actors: 2
    • Video Length: ~5 sec

• Scene 3

  • In this scene, there will be a couple shots of people moving still pictures (in frame) toward the camera; the still image is the first frame for the next video. The last picture frame will hold the image of the beginning of Scene 4.
    • Number of Actors: 3
    • Video Length: ~5 sec

• Scene 4

  • Scene 4 will be a single person holding a frame in a scene with people acting in the background. The people in the background will be walking toward each other and when they are inside the picture frame it will move to a still image. When the still image is removed, the people are actually fighting.
    • Number of Actors: 3
    • Video Length: ~5 sec

Scene 5	Scene 6	Scene 7	Scene 8

• Scene 5

  • This is the scene from the HP PictureBook Ad where the guy is in front of the printer putting picture frames over his head. We intend to do something similar to this, but when he puts one of the frames over his head he then has a mask on.
    • Number of Actors: 1
    • Video Length: ~5 sec

• Scene 6

  • Scene 6 is the party scene; here, an arm will reach out from the side of the screen to pull still images from the video. When these pictures are pulled from the scene, the person in the video now has something wrong with them; for example, pulling a picture from someone's face may result in them now having a black eye.
    • Number of Actors: 3
    • Video Length: ~5 sec

• Scene 7

  • Kid running up hill, w/frames catching his movement up the hill in pictures, but he runs into the last one and falls down (but the picture is of him running into and starting to fall).
    • Number of Actors: 7
    • Video Length: ~5 sec

• Scene 8

  • People from Scene 1 are holding picture frames which consist of pieces of the final Geico-like Logo; the people put these picture frames together and the camera zooms in on the result – a picture of the Geico Logo (from TV Ad) – while saying "The pictures are real, but they won't save you any money on your car insurance!" When the Geico Logo is full-screen (with phone number and everything) say "Geico... 15 minutes could save you 15 percent or more."
    • Number of Actors: 10
    • Video Length: ~10 sec

The Input, The Output, The Layers

• Scene 1

  • The Input:
    • A footage with ten actors standing in rows in a courtyard. Each of them holds a frame, and later they exchange the frame each other.
    • Several still images extracted from the footage.
  • The Output:
    • A video with ten people stand in rows. They exchange boards with the still images of their heads.
  • The Layers:
    • Layer1: extracted still images.
    • Layer2: footage with people.
    • The still images of heads will be put on top of the footage and fill exactly into the regions within the frames.

• Scene 2

  • The Input:
    • A footage with a woman sitting on a bench with green screen as background, someone keeps handing blue cards to her.
    • A footage with a man jumping up and down on a trampoline. Near the end, the trampoline is removed and the man fall down onto the ground.
    • Several still images extracted from the footage with man jumping.
  • The Output:
    • A video with a woman sitting on the bench has her hand reach in to the air and pull out cards with images of the man jumping up and down in the background. Last time, instead of pulling out the jumping man, she pulls out an image of the trampoline, and the trampoline looks indeed been removed at the same time. The man falls down onto the ground as result.
  • The Layers:
    • Layer1: footage with a man jumping up and down
    • Layer2: footage with a woman sitting on a bench
    • Layer3: still images of the jumping man & the trampoline
    • The footage with woman sitting on a bench will be put on top of the man jumping layer and the regions of green screen will be replaced by the layer beneath it. On top of those, still images of the jumping man and the trampoline cover and replace the regions of blue cards.

• Scene 3

  • The Input:
    • Several videos of actor holding picture frame and moving toward camera
    • First frame of each "next" video to put on the picture frame for the "previous" video
  • The Output:
    • Single video of multiple actors walking toward camera (one at a time) holding a picture of the first frame of the next video.
  • The Layers:
    • Layer1: Footage of the actors walking toward the camera
    • Layer2: still images of the next video placed in the picture frame in the previous video
    • The last video will have a picture frame with the first frame of Scene 4

• Scene 4

  • The Input:
    • Main footage of person holding picture frame in front of moving scene, with people walking behind him
    • Still image extracted from the video
  • The Output:
    • Single video of person holding a picture frame where the people in the background turn into the still image in the picture frame. When the picture frame is removed, the people are actually fighting.
  • The Layers:
    • Layer1: Main footage will be used as the background layer
    • Layer2: Still image when the people are both in the picture frame will be layered on top of the picture frame.

• Scene 5

  • A footage with a man standing by a desk who keeps putting frames over his head
  • A footage with the same man in the same position, but with a mask on
  • Some still images extracted from the videos
  • The Output:
    • A video. We are going to composite the previously extracted still images to fill the regions within the frames. Footage one is going to switch into footage two to make the effect that the man gets a mask after putting a frame over his head.
  • The Layers:
    • Layer 1: The two footages
    • Layer 2: Still images
    • The two footages are going to be connected without overlap in time, hence they can be put on the same layer. We are going to fill the regions within the picture frames with still images. The region within the last picture frame which is put over the first man's head will be filled by the image cropped from the first time frame of the second video.

• Scene 6

  • The Input:
    • Background video of party scene
    • Foreground video of actor's hand reaching from side of screen to grab pictures (filmed in front of green screen)
    • Blue card in person's hand which will be the picture from the scene
  • The Output:
    • Single video of person reaching out to grab pictures of the party scene, where the underneath video will have the person altered in some way (e.g., black eye).
  • The Layers:
    • Layer1: Background video will be used to capture small pictures which actor will be grabbing from the scene
    • Layer2: Foreground video of actor's hand will be chromakeyed to take out the green background and the blue card will hold the still image (using position markers)
    • Layer3: Blue card in person's hand will hold the still image (using position markers)

• Scene 7

  • The Input:
    • A footage with a camera zooms out. The camera moves through the frames which held by pairs of people. A kid runs forward to the camera and finally has his head hit the last frame. Several still images extracted from the footage mentioned above.
  • The Output:
    • A video looks just like the input footage only the regions in the frames are replace by still images.
  • The Layers:
    • Layer1: the footage.
    • Layer2,3,4...: still images (the number depends on how many frames we decided to zoom out through)

• Scene 8

  • The Input:
    • A footage with people from Scene 1 holding 9 white blank boards. They hold the boards in the way that the boards align one by one and form a 3x3 plane. They later move the boards apart from each other.
    • Still images of Geico logo.
  • The Output:
    • Each person in scene 1 holds a board with a piece(1/9) of the Geico logo image on it. They then hold and aligned the boards to form the Geico logo. Then the camera zoom-in until the Geico logo fill the whole screen.
  • The Layers:
    • Layer1: extracted still images of Geico logo.
    • Layer2: footage with people.
    • The still images of Geico logo will be put on top of the footage and fill exactly into the regions of the frames and within the frames. The footage will be used in reverse order.

Who does what

• Ahren:
  • Storyline/Storyboards
  • Shot Planning
  • Filming
  • OpenCV coding
  • Video Editing
  • Casting
• Ang:
  • Storyline/Storyboards
  • Shot Planning
  • Filming
  • Video Editing
  • Casting
  • Acting
• Ed:
  • Shot Planning
  • Filming
  • OpenCV coding
  • Video Editing
  • Casting
  • Props

Timeline

• March 1: Proposal Due
• March 13: Experimental Footage and Effect
• March 15: Midterm Report Due
• March 25: Footage Complete
• March 29: Project Update 1
• April 3: Effect Coding Complete
• April 5: Project Update 2
• April 10: Final Editing Complete
• April 12: Project Update 3
• April 14: Internal Screening
• April 28: External Screening
• May 2: Final Report Due