Abstract: Disclosed herein are image-based object recognition method and system by and in which a learning server performs image-based object recognition based on the learning of environment variable data. The image-based object recognition method includes: receiving an image acquired through at least one camera, and segmenting the image on a per-frame basis; primarily learning labeling results for one or more objects in the image segmented on a per-frame basis; performing primary reasoning by performing object detection in the image through a model obtained as a result of the primary learning; performing data labeling based on the results of the primary reasoning, and performing secondary learning with weights allocated to respective boxes obtained by the primary reasoning and estimated as object regions; and finally detecting one or more objects in the image through a model generated as a result of the secondary learning.

Abstract: A color conversion panel includes a first color filter and a second color filter that are disposed on a substrate, a low refractive index layer disposed on the substrate, the first color filter, and the second color filter, the low refractive index layer including at least one of a first blue pigment and a first blue dye, a first color conversion layer overlapping the first color filter and including a semiconductor nanocrystal, a second color conversion layer overlapping the second color filter and including a semiconductor nanocrystal, and a transmissive layer that overlaps the low refractive index layer.

Abstract: Provided is a method for purifying an anti-4-1BB/anti-HER2 bispecific antibody, the method of which includes carrying out affinity chromatography with a sodium acetate buffer containing a certain inorganic salt as an elution buffer. The purification method of the presently claimed subject matter increases the elution of an antibody in the intact form, thereby being able to provide an anti-4-1BB/anti-HER2 bispecific antibody in the intact form in high purity and high yield.

Abstract: Disclosed herein are image-based object recognition method and system by and in which a learning server performs image-based object recognition based on the learning of environment variable data. The image-based object recognition method includes: receiving an image acquired through at least one camera, and segmenting the image on a per-frame basis; primarily learning labeling results for one or more objects in the image segmented on a per-frame basis; performing primary reasoning by performing object detection in the image through a model obtained as a result of the primary learning; performing data labeling based on the results of the primary reasoning, and performing secondary learning with weights allocated to respective boxes obtained by the primary reasoning and estimated as object regions; and finally detecting one or more objects in the image through a model generated as a result of the secondary learning.

Abstract: A method for manufacturing a vehicle knuckle is provided. The method includes preparing a die for forming the vehicle knuckle, locating at least one busing in the die, filling the die with carbon chip materials, hot press forming the carbon chip materials filled in the die at high temperature and high pressure environment, separating the formed vehicle knuckle from the die, and trimming and removing surplus materials attached to the separated vehicle knuckle.

Abstract: According to one embodiment of the present disclosure, a vehicle knuckle is provided. The vehicle knuckle according to one embodiment of the present disclosure may comprise a knuckle body provided with a wheel bearing installation hole configured to install a wheel bearing, and flanges extending from the knuckle body and used to connect the vehicle knuckle to other parts of a vehicle. The knuckle body and the flanges may be integrally formed of carbon chip materials.