Inner Access Consulting - Gary Feierbach |
HomeBioNewsExperienceArtificial IntelligenceRoboticsCareer HighlightsPublicationsPatentsLinkedIn PageWordPress BlogMy Alter Ego
|
Computer Hardware and Software Task ListTask: Speech Recognition and SynthesisPurpose Speech to text, text to speech, Voice location Hardware Raspberry Pi 3B computer 1 USB microphone 1 1W Audio Amplifier 1 3 in Speaker 1 microSD - class 10 card 32Gb (Perhaps a couple more for backup) Software Resources Jasper's Speech recognition and voice control https://jasperproject.github.io Oscar Liang's Voice to text for control and text to speech synthesis https://oscarliang.com/raspberry-pi-voice-recognition-works-like-siri/ Steven Hickson's Speech recognition and voice control. https://stevenhickson.blogspot.in/2013/04/voice-control-on-raspberry-pi.html Task: USBNET between Raspberry PIs Purpose USB high bandwidth communication between Raspberry PIs Hardware Belkin USB Direct connect Software> USB Driver for USBNET requires linux that supports USBNET Resources USB in a nutshell GNU/Linux USBNET Driver Framework Task: Right (Dominant) Eye Vision Purpose Box and recognize faces, moving objects Hardware Raspberry Pi 3B computer 1 Raspberry Pi Camera module V2 (8 mega pixels 1080P) Software Resources Raspberry tutorials on Computer Vision https://www.intorobotics.com/20-hand-picked-raspberry-pi-tutorials-in-computer-vision/ Task: Left (Slave) Eye Vision Purpose Align with right eye point of interest and estimate distance, also note positions of moving objects Reconstruct space volume and objects in volume. Hardware Raspberry Pi 3B computer 1 Raspberry Pi Camera module V2 (8 mega pixels 1080P) Software Measure object overlap with right central vision and minimize with eye movement. Read back measure and track. Report. Task: Sensor integration Purpose Collect information from all sensors, interpret and act on same. All information will be collected via USB buses Build map of space. Navigate Make it all displayable Hardware Raspberry Pi 3B Computer 1 top of body 3D accelerometers and 3D gyros in top of body. (LSM330DLC) 1 GPS Tracker in top of head near surface. (Trimble Copernicus II) Conversational Intelligence (language understanding) People recognition/history Knowledge base Hierarchical command base Short term memory machine (learns temporarily and tags information for long term memory training) Build sketch of space and objects therein for navigation. Display all inputs on large screen including selected parts on vision inputs Since this is the main brains of the robot more compute power may be necessary such as the Epiphany 1024 processor. It could also be used for night training sessions as well. Task: Leg motion processor This processor will contain a large number of pre-trained macros used for walking, sitting standing, steering through objects, etc. Hardware Raspberry Pi 3B Computer 8 3D accelerometers and 3D gyros in all limb segments. (LSM330DLC) Task: Arm and hand motion processor Putpose: This processor will contain pre-trained macros for coordinated arm, hand and finger motions, for picking up, handling, operating a device, using an object, shaking hands, sign language, etc. Hardware Raspberry Pi 3BComputer |