Research

IN THE PROCESS, WE SPEED UP IMAGE GENERATION BY UP TO 2X AND REDUCE MEMORY CONSUMPTION BY UP TO 5. 6X.

WE INTRODUCE GPT-NEOX-20B, A 20 BILLION PARAMETER AUTOREGRESSIVE LANGUAGE MODEL TRAINED ON THE PILE, WHOSE WEIGHTS WILL BE MADE FREELY AND OPENLY AVAILABLE TO THE PUBLIC THROUGH A PERMISSIVE LICENSE.

TO EFFECTIVELY FUSE LANGUAGE AND VISION MODALITIES, WE CONCEPTUALLY DIVIDE A CLOSED-SET DETECTOR INTO THREE PHASES AND PROPOSE A TIGHT FUSION SOLUTION, WHICH INCLUDES A FEATURE ENHANCER, A LANGUAGE-GUIDED QUERY SELECTION, AND A CROSS-MODALITY DER FOR CROSS-MODALITY FUSION.

IN THIS NOTE, WE EXPLORE INFERENCE-TIME ALIGNMENT THROUGH IN-CONTEXT LEARNING.

ON THE ONE HAND, EACH QUERY IS GENERATED BASED ON 2D LANE-AWARE FEATURES AND ADOPTS A HYBRID EMBEDDING TO ENHANCE THE LANE INFORMATION.

WE PRESENT C-EVAL, THE FIRST COMPREHENSIVE CHINESE EVALUATION SUITE DESIGNED TO ASSESS ADVANCED KNOWLEDGE AND REASONING ABILITIES OF FOUNDATION MODELS IN A CHINESE CONTEXT.

BASED ON THIS IDEA, WE PRESENT MEMORY-AND-ANTICIPATION TRANSFORMER (MAT), A MEMORY-ANTICIPATION-BASED APPROACH, TO ADDRESS THE ONLINE ACTION DETECTION AND ANTICIPATION TASKS.

SINGLE-STAGE TEXT-TO-SPEECH MODELS HAVE BEEN ACTIVELY STUDIED RECENTLY, AND THEIR RESULTS HAVE OUTPERFORMED TWO-STAGE PIPELINE SYSTEMS.

UNLIKE TRADITIONAL MULTI-VIEW STEREO APPROACHES, THE NEURAL IMPLICIT SURFACE-BASED METHODS LEVERAGE NEURAL NETWORKS TO REPRESENT 3D SCENES AS SIGNED DISTANCE FUNCTIONS (SDFS).

FINE-TUNING LANGUAGE MODELS (LMS) HAS YIELDED SUCCESS ON DIVERSE DOWNSTREAM TASKS, BUT AS LMS GROW IN SIZE, BACKPROPAGATION REQUIRES A PROHIBITIVELY LARGE AMOUNT OF MEMORY.

IN OUR EXPERIMENTS, WE TREAT DISCRETE ACOUSTIC S AS TEXTUAL DATA AND TRAIN A MASKED LANGUAGE MODEL USING A CLOZE-LIKE METHODOLOGY, ULTIMATELY DERIVING HIGH-QUALITY AUDIO REPRESENTATIONS.

IN THIS , WE EXPLORE THE POSSIBILITY OF BOOSTING DEEP 3D POINT CLOUD ENRS BY TRANSFERRING VISUAL KNOWLEDGE EXTRACTED FROM DEEP 2D IMAGE ENRS UNDER A STANDARD TEACHER-STUDENT DISTILLATION WORKFLOW.

APPLICATIONS BUILT ON TOP OF LARGE LANGUAGE MODELS (LLMS) SUCH AS GPT-4 REPRESENT A REVOLUTION IN AI DUE TO THEIR HUMAN-LEVEL CAPABILITIES IN NATURAL LANGUAGE PROCESSING.

TEXT EVALUATION HAS HISTORICALLY POSED SIGNIFICANT CHALLENGES, OFTEN DEMANDING SUBSTANTIAL LABOR AND TIME COST.

TRACKING 3D OBJECTS ACCURATELY AND CONSISTENTLY IS CRUCIAL FOR AUTONOMOUS VEHICLES, ENABLING MORE RELIABLE DOWNSTREAM TASKS SUCH AS TRAJECTORY PREDICTION AND MOTION PLANNING.

WE OBSERVE THAT THE INEFFICIENCY IS DUE TO SUBOPTIMAL WORK PARTITIONING BETWEEN DIFFERENT THREAD BLOCKS AND WARPS ON THE GPU, CAUSING EITHER LOW-OCCUPANCY OR UNNECESSARY SHARED MEMORY READS/WRITES.

THE MISUSE OF LARGE LANGUAGE MODELS (LLMS) HAS GARNERED SIGNIFICANT ATTENTION FROM THE GENERAL PUBLIC AND LLM VENDORS.

MS3D++ PROVIDES A STRAIGHTFORWARD APPROACH TO DOMAIN ADAPTATION BY GENERATING HIGH-QUALITY PSEUDO-LABELS, ENABLING THE ADAPTATION OF 3D DETECTORS TO A DIVERSE RANGE OF LIDAR TYPES, REGARDLESS OF THEIR DENSITY.

LCENSEMBLE IS A HIGH-PERFORMING, SCALABLE AND USER-FRIENDLY PYTHON PACKAGE FOR THE GENERAL TASKS OF CLASSIFICATION AND REGRESSION.

OUR SYSTEM DISENTANGLES THIS OBJECTIVE INTO THREE SEQUENTIAL TASKS: (1) FACE VIDEO GENERATION WITH A CANONICAL EXPRESSION; (2) AUDIO-DRIVEN LIP-SYNC; AND (3) FACE ENHANCEMENT FOR IMPROVING PHOTO-REALISM.