CDNN software framework, in conjunction with the CEVA-XM4 imaging and vison DSP, enables: - Real-time object recognition and vision analytics - Lowest power deep learning solution for embedded systems: 30x lower power and 3x faster processing when compared to leading GPU-based systems - 15x average memory bandwidth reduction compared to typical neural network implementations - Automatic conversion from offline pre-trained networks to real-time embedded-ready networks - Flexibility to support various neural network structures, including any number and type of layers[Continue reading →]
MOUNTAIN VIEW, Calif., Feb. 14, 2014 -- CEVA, Inc. (NASDAQ: CEVA), the leading licensor of silicon intellectual property (SIP) platform solutions and DSP cores, and Sensory, Inc., a pioneer in speech technologies for consumer products, today announced the availability of power-optimized solutions aimed at bringing advanced voice activation capabilities to a new wave of mobile and sensor-enabled devices.
The solutions integrate Sensory's market-leading TrulyHandsfree™ Voice Control Version 3.0 optimized for the CEVA-TeakLite-4 V2 DSP architecture, to address a variety of consumer applications that demand ultra-low power consumption for "always on, always listening" voice control functions. For example, the solution consumes less than 20 microwatts for voice triggering in 28nm, making it ideal for always-listening wearable devices such as smart watches and glasses, as well as integration into sensor hub solutions in smartphones. The companies will showcase the latest TrulyHandsfree features optimized for the enhanced CEVA TeakLite-4 V2 architecture at the 2014 Mobile World Congress in the CEVA booth (Hall 6, Stand A50).[Continue reading →]
MOUNTAIN VIEW, Calif., and OULU, Finland, Feb. 14, 2014 -- CEVA, Inc. (NASDAQ: CEVA), the leading licensor of silicon intellectual property (SIP) platform solutions and DSP cores, and Visidon, Ltd., a developer of industry-leading facial recognition technologies for mobile and embedded platforms, today announced that the companies have partnered to develop an ultra-low power, 'always-on' face activation solution targeting mobile devices. The solution combines Visidon's face activation software with the CEVA-TeakLite-4 sensor fusion platform to enable highly power efficient context aware applications that can activate a smartphone, tablet or any other camera-enabled device upon detecting the presence of a user's face. CEVA and Visidon will demonstrate this solution at the upcoming Mobile World Congress in CEVA's booth (Hall 6, stand A50).
Face activation technology is set to become a key component of low-power sensor fusion applications, where a device can remain in standby mode while awaiting activation via visual presence or voice commands. In mobile devices, implementing these 'always-on' technologies utilizing absolute minimal power consumption is crucial, resulting in longer battery life for the device. By utilizing the CEVA-TeakLite-4 sensor fusion platform rather than the main device CPU to implement Visidon's face activation software consumes less than 1mA at a 28nm process. In addition, by only using a fraction of the CEVA-TeakLite-4 available MHz, there is ample headroom on the processor to run additional applications in parallel, such as 'always-listening' voice activation.[Continue reading →]
- Feature-rich RivieraWaves Bluetooth Smart 4.2 IP incorporates advanced features including data length extension, secure connection, enhanced privacy and scatternet operation - Widely adopted by multiple licensees for various IoT, wearable and smart home applications - Complements CEVA's portfolio of RivieraWaves Bluetooth Smart Ready 4.2 IP ("Dual Mode") and RivieraWaves Wi-Fi 802.11a/b/g/n/ac IP platforms[Continue reading →]
New CEVA-XM4 imaging and vision IP takes embedded vision one step closer to human vision, enabling: - Real-time 3D depth map and point cloud generation - Deep learning and neural network algorithms for object recognition and context awareness - Computational photography for image enhancement including zoom, image stabilization, noise reduction and improved low-light capabilities[Continue reading →]
Latest 3GPP VoLTE codec offers superb voice clarity and audio quality, requiring substantially higher DSP performance; Implementation for the CEVA-TeakLite family of DSPs offers best power and performance solution for smartphones and other mobile computing devices[Continue reading →]
SANTA CLARA, Calif. Linley Mobile Conference -- CEVA, Inc. (NASDAQ: CEVA), the leading licensor of silicon intellectual property (SIP) platform solutions and DSP cores today announced that the CEVA-TeakLite-4 DSP is now handling the processing load of Bluetooth, in addition to audio, voice and sensing technologies, thereby dramatically lowering the cost, complexity and power consumption of chip designs targeting smartphones, the Internet of Things (IoT), wearables and wireless audio devices. Leveraging new instructions and interfaces added to the recently introduced multifunctional CEVA-TeakLite-4 v2 architecture, the DSP now runs CEVA-Bluetooth connectivity (Classic or Low Energy) in addition to a broad range of audio and voice software packages, 'always-on' user interface (UI) functions such as voice triggering and face activation, and a full suite of sensor-fusion capabilities, all on a single core. CEVA is presenting this multifunctional DSP today at the Linley Mobile Conference, in Santa Clara, California.
The Internet of Things comprises a multitude of devices, technologies and form factors, with many use cases and requirements. The CEVA-TeakLite-4 specifically targets user-centric IoT devices, where natural user interface, audio playback and voice communication represent key attributes of the device. This can include for example, voice activation, face triggering and other 'always-on' functionality in a smartphone, smart watch, smart home controller or wireless speakers. The ultra-low power nature of the CEVA-TeakLite-4 DSP ensures that these 'always-on' features consume minimal battery life. All of this functionality can run concurrently on the DSP without the need for a host CPU, reducing the die size and lowering power consumption of the overall device. Illustrating this, a real-life use case implementing Bluetooth Low Energy, always-on UI and sensor fusion on the CEVA-TeakLite-4 DSP requires less than 150K gates and consumes less than 150uW when implemented in a 28nm process.[Continue reading →]
MOUNTAIN VIEW, Calif. -- CEVA, Inc. (NASDAQ: CEVA), the leading licensor of DSP-based IP platforms for vision, audio, communications and connectivity, today announced that Novatek Microelectronics Corp., a leading fabless chip design company, has selected the CEVA-MM3101 imaging and vision DSP to power its next-generation System-on-Chips (SoC) targeting the surveillance, action camera and automotive markets. Novatek will integrate the programmable CEVA-MM3101 into their SoC designs to add powerful computer vision capabilities including scene analysis, machine vision, depth mapping and object detection in a flexible and power efficient manner.
"Integrating advanced vision processing capabilities with our imaging SoCs allows our customers to design products offering a new level of interactivity with the real world and show real product differentiation," Tommy Chen, Vice President at Novatek. "The CEVA-MM3101 imaging and vision DSP provides the processing horsepower while extending the device battery life and flexibility to implement innovative, intelligent imaging algorithms for any end application."[Continue reading →]