Publications

• Multi-Bank On-chip Memory Management Techniques for CNN Accelerators
• S3NAS: Fast NPU­aware Neural Architecture Search Methodology
• Scheduling of Deep Learning Applications onto Heterogeneous Processors in an Embedded Device
• C-GOOD: C-code Generation Framework for Optimized On-device Deep Learning
• Joint Optimization of Speed, Accuracy, and Energy for Embedded Image Recognition Systems
• Architectures and Algorithms for User Customization of CNNs
• Work-­in-­Progress: Incremental Training of CNNs for User Customization
• FIFA: A Kernel­Level Fault Injection Framework for ARM­based Embedded Linux System
• A space- and energy-efficient code compression/decompression technique for coarse-grained reconfigurable architectures
• Real-time co-scheduling of multiple dataflow graphs on multi-processor systems
• Task-Processor ILP Scheduling for CNN on Heterogeneous Computing environment Embedded Device
• Co-scheduling Technique of Dataflow Applications with Shared Processor Allocation
• Co-Scheduling of Multiple Concurrent SDF Applications with Arbitrary Periods and Offsets