SOLUTIONS
SOLUTIONS
Here is what the final visual would look like when any digital content is played on a device with the system installed
Here is what the final visual would look like when any digital content is played on a device with the system installed. The solution includes few of the most refined versions of the filters. Having this filter incorporated in videos opens up a whole new way of watching movies. This is an attempt to create a less intrusive yet extremely powerful and compelling superimposed layers that is seamlessly integrated with the original visual. This is a project made specifically for a cause and to help people with disabilities enjoy cinema. It is for us to acknowledge that art must be created in such a way that it is appreciated by everyone. The sole purpose of having such a filter is to provide the opportunity for all sorts of viewers to have a great movie experience.
Breakdown of filter
This video clipping shows the complete breakdown of the filter. Bringing in all the layers together to form a well structured and sandwiched visual system.
Breakdown of filter
Motion piece of just the 2D filter with all visual attributes included. Different themes were tested at first. Darker theme is used for presentation purposes. Users can easily toggle between a variety of layouts based on their personal preferences. This simulation like most other earlier versions shown in the process page demonstrates the use of shapes, beats, rhythm, color and visualized audio wave-forms.
Modes of display
When a movie is being played on a device such a TV or a laptop ,the visual filter augmented over the original video will occupy the entire screen. You can switch between two modes - Overlay or Icon
Use of Overlay
Layered and superimposed computer generated visuals , that include periodically changing clusters driven by python fields , particle simulations and overlapped shapes with colors
Use of Icons
This is an approach that incorporates the use of simple and minimalist designs (shapes in combination with changing colors) to communicate emotion and context
How does it look on a device ?
The filters occupy the entire screen . Aspect ratios can be changed at any time and even the themes can be decided according to personal preferences. Changes made based on personal preferences would not alter the grammar of the visual system. Hue-Saturation values for color and particle speeds, are at the end of the day relative. When you are playing the video , the user will be prompted to select a viewing mode.
Simulated visuals for different scenes
The visual filter changes based on the scenes .Here are a few examples of how the filter would look like for different scenes. Each of these clippings would have different emotional content , context and dialogues associated with the scene.
Primary Emotions
Filters simulated for a specific emotion that seems obvious and direct can be created very easily, since the visual interpretation will be straightforward. The tracker will detect facial expressions and will create a layer for context based on just the obvious emotions associated with the looks on the subject's face and background score in conjunction with the script. The context of the scene might not be hard to decode by AI in such cases.
The whole point of have a multi-layered approach is to smartly flood the composition and provide the audience with enough visual information on screen so that there wont be any dependence on audio at all. In the case of having just a primary emotion , there will not be a necessity to utilize all layers ,and therefore the bottom most layer reserved for providing content of the scene will merely mimic the layers on top. There won't be a mixture of colors in such compositions. Single color corresponds to a single emotion . The intensities of emotion and context will be decided by what sort of shapes are being projected on the screen.
No matter how much a visual is manipulated ,there is always going to be a lot of room for subjectivity. The movies that we watch today are already manipulated to a great extent, but we don't often tend to realize it since we are used to watching digital content in a way that would put us all in a position where we will watch the film from the director's point of view.
Some of the clippings shown below have different intensity levels but all of them have visual context provided based on only script and facial expressions.
Circle - Degree I ( Feeling Low )
Blue - Depressing
​
Tempo I - Pattern driven by Python fields
Hexagon - Degree II ( Physical / Tensed )
​
Red - Rage
​
Tempo II - Pattern driven by Python fields with moderate frequency ( Background )
Square - Degree III ( Excited )
​
Yellow - Warmth / Love
​
Tempo III - Pattern driven by Python fields with high frequency ( Background )
Square - Degree III ( Hysterical / Horrifying )
​
Green - Creepy / Vulgar
​
Tempo IV - Pattern driven by Python fields with maximum frequency ( Background )
Complex Emotions
There are scenarios where we might find the underlying context to be different from what is evidently shown . In such cases , the system has to track more parameters other than just the dialogs ( audio ) and facial expressions ( video - facial tracking ) .
For the system to understand the whole scene and develop a visual system with multiple layers , it has to read all audio files including background score , camera movements and figure out the mood of the scene based on the script.
One must acknowledge that even scenes with gentle camera work that are extremely quiet can be overwhelming for the viewer. Therefore it is vital for the system go through the script line by line and create the intensity levels with the help of AI ( Artificial Intelligence ) ,a lot of permutations must be taken into consideration if the system has to provide context in form of visual information
A video filter of a couple having a serious conversation .What looks evident is the love they are showing each other while talking. The context is probably something depressing and serious ,which is not the same as what we see on screen. When the audio is removed ,you lose the sensation and feeling created by the background music which was originally composed to enhance the cinematic feel. Subtler emotions utilize shapes like circles and hexagons ,since its less intense due to its roundness
A similar clipping with a more intense context ,would require the use of shapes such as squares. The circles masking the faces of the two characters have a mild rippling effect which is controlled by dialogs ( audio waveform excluding background music and sound ). In most scenes , the character may or may not look directly into the camera but the dialogue might still be delivered .Therefore , the shapes exist throughout once the character makes an entrance. Without audio it will be hard to follow along, when the character turns away from the camera ( even the slightest of sounds made by the character matter , doesn't have to be lengthy dialogues necessarily.